1. Regulatory Review of Higher Phase Clinical
Trials
ICMR sponsored and Sat Kaival College of Pharmacy organized
National Symposium on EMERGING TRENDS IN CLINICAL RESEARCH
29th February, 2008
Dr. Bhaswat S. Chakraborty
Sr. VP, R&D, Cadila Pharmaceuticals
1

2. Order of the Topics
Clinical Trials
– Regulatory phases 1,2,3 & 4
Review of Clinical Trials
Data Requirements
Study Design Considerations
Controls
– Placebo, Active, NI
Assay Sensitivity, Multiplicity of Analyses
Interim Analysis
Intent to Treat Analysis
Final Analysis
Conclusions
2

3. 3

4. Investigational New Drug
The pharmaceutical industry sometimes provides advice to
the FDA prior to submission of an IND
Sponsors, research institutions, and other organizations
that take responsibility for developing a drug must show
the FDA results of preclinical testing they've done in
laboratory animals and what they propose to do for human
testing
At this stage, the FDA decides whether it is reasonably
safe for the company to move forward with testing the
drug in humans
4

5. Phase 1
Phase 1 studies are usually conducted in healthy
volunteers
The goal here is to determine what the drug's most
frequent side effects are and, often, how the drug
is metabolized and excreted
The number of subjects typically ranges from 20
to 100
5

6. Early Phase Clinical Trials
6

7. Phase 2
Phase 2 studies begin if Phase 1 studies don't reveal unacceptable toxicity
While the emphasis in Phase 1 is on safety, the emphasis in Phase 2 is on
effectiveness
This phase aims to obtain preliminary data on whether the drug works in
people who have a certain disease or condition
For controlled trials, patients receiving the drug are compared with similar
patients receiving a different treatment--usually an inactive substance
(placebo), or a different drug
Safety continues to be evaluated, and short-term side effects are studied.
Typically, the number of subjects in Phase 2 studies ranges from a few
dozen to about 300
7

8. Phase 3
At the end of Phase 2, the FDA and sponsors try to come to an
agreement on how the large-scale studies in Phase 3 should be done
How often the FDA meets with a sponsor varies, but this is one of two
most common meeting points prior to submission of a new drug
application
The other most common time is pre-NDA, right before a new drug
application is submitted
Phase 3 studies begin if evidence of effectiveness is shown in Phase 2
These studies gather more information about safety and effectiveness,
studying different populations and different dosages and using the drug
in combination with other drugs
The number of subjects usually ranges from several hundred to about
3,000 people
8

9. 9

10. Phase 4
Postmarketing study commitments are called
Phase 4 commitments
– studies required of or agreed to by a sponsor
that are conducted after the FDA has approved
a product for marketing
The FDA uses postmarketing study commitments
to gather additional information about a product's
safety, efficacy, or optimal use
10

11. 11

12. Examples of Regulatory Approvals: USFDA
12

13. Goals of NDA Review
Whether the drug is safe and effective in its proposed
use(s), and whether the benefits of the drug outweigh the
risks.
Whether the drug's proposed labeling (package insert) is
appropriate, and what it should contain.
Whether the methods used in manufacturing the drug and
the controls used to maintain the drug's quality are
adequate to preserve the drug's identity, strength, quality,
and purity.
13

14. Drug Review Steps
1. Preclinical (animal) testing.
2. An investigational new drug application (IND) outlines what the sponsor of a new
drug proposes for human testing in clinical trials.
3. Phase 1 studies (typically involve 20 to 80 people).
4. Phase 2 studies (typically involve a few dozen to about 300 people).
5. Phase 3 studies (typically involve several hundred to about 3,000 people).
6. The pre-NDA period, just before a new drug application (NDA) is submitted. A
common time for the FDA and drug sponsors to meet.
7. Submission of an NDA is the formal step asking the FDA to consider a drug for
marketing approval.
8. After an NDA is received, the FDA has 60 days to decide whether to file it so it can
be reviewed.
9. If the FDA files the NDA, an FDA review team is assigned to evaluate the sponsor's
research on the drug's safety and effectiveness.
10. The FDA reviews information that goes on a drug's professional labeling (information
on how to use the drug).
11. The FDA inspects the facilities where the drug will be manufactured as part of the
approval process.
12. FDA reviewers will approve the application or find it either "approvable" or "not
approvable."
14

15. Review Process
Once a new drug application is filed
– an FDA review team evaluates whether the studies the sponsor submitted show that
the drug is safe and effective for its proposed use
Team consists of medical doctors, chemists, statisticians, microbiologists,
pharmacologists, and other experts
No drug is absolutely safe; all drugs have side effects
– "Safe" in this sense above means that the benefits of the drug appear to outweigh
the risks.
The review team
– analyzes study results
– looks for possible issues with the application
e.g., weaknesses of the study design or analyses
may agree with the sponsor's results and conclusions, or may need additional information
to make a decision
Each reviewer prepares a written evaluation containing conclusions and
recommendations about the application
These evaluations are then considered by team leaders, division directors, and office
directors, depending on the type of application
15

16. Clinical Trials: Testing Medical Products in Humans
Clinical studies, test potential treatments in human
volunteers to see whether they should be approved for
wider use in the general population
– A treatment could be a drug, medical device, or
biologic, such as a vaccine, blood product, or gene
therapy
– A new treatment may or may not be “better”
– Complete and accurate research
– Protection and well being of participants
Ethics, consent, audit
– Documentation
16

17. Clinical Trials..
Drug studies in humans can begin only after an
IND is reviewed by the FDA and a local
institutional review board (IRB)
The board is a panel of scientists and non-
scientists in hospitals and research institutions that
oversees clinical research
17

18. Institutional Review
IRBs approve the clinical trial protocols
– the type of people who may participate in the clinical
trial
– the schedule of tests and procedures
– the medications and dosages to be studied
– the length of the study
– the study's objectives
– other details
IRBs make sure the study is
– acceptable
– participants have given consent
– Participants are fully informed of their risks
– researchers take appropriate steps to protect patients from
harm 18

19. Data
There are legal and ethical reasons for reporting all
relevant data collected during the drug development
process
Some reporting strategies already exist in the 1988
Guidelines, ICH E3 and E9
Electronic Submissions and desktop review capabilities
will help all of us make better use of clinical data in
NDA’s
There may be better strategies and these should be
considered
19

20. Source: MDS, New Zealand
20

21. What does FDA Look for?
FDA approves a drug application based on
– Substantial evidence of efficacy & safety from
“adequate and well-controlled investigations”
– A valid comparison to a control
– Quantitative assessment of the drug’s effect
(21 CFR 314.126.)
The design of cancer trials intended to support
drug approval is very important
21

22. Adequate and Well-Controlled Studies
Because the course of most diseases is variable, you need a control
group, a group treated just like the test group, except that they don’t
get the drug, to distinguish the effect of the drug from spontaneous
change, placebo effect, observer expectations
21 CFR 314.126 describes the following controls
– Placebo
– No treatment
– Dose response
– Active control
Superiority of non-inferiority
– Historical
Placebo, dose response or superiority are usually convincing studies
22

23. Adequate and Well-Controlled Studies..
Minimization of Bias: a unidirectional tilt favoring one
group, i.e., a non-random difference in how test and
control group are selected, treated, observed, and analyzed
– These are the 4 main places bias can enter
Remedies:
– Blinding (patient and observer bias)
– Randomization (treatment and control start out equal)
– Careful specification of procedures and analyzes in a
protocol to avoid
Choosing the most favorable analysis out of many (bias)
Having so many analyses that one is favorable by chance
(multiplicity)
Source: RJ Temple, US FDA, Unapproved Drugs Workshop January 2007
23

24. Design Concepts
Difference in Clinical Efficacy (Є)
Non-Inferiority
Superiority
+δ
0
Equivalence
-δ
Inferiority
Non-Superiority
Equality δ = Meaningful Difference
24

25. Purposes of Active Trials
The purpose of an active control trial could
be to demonstrate that a new experimental
treatment is either
superior to the control
equivalent to the control, or
non-inferior to the control
superior to a virtual placebo
25

26. Study Design: Approaches
Randomised Controlled Trials (RCT) most preferred approach
– Demonstrating superiority of the new therapy
Other approaches
– Single arm studies (e.g., Phase II)
e.g., when many complete responses were observed or when
toxicity was minimal or modest
– Equivalence Trials
– No Treatment or Placebo Control Studies
– Isolating Drug Effect in Combinations
– Studies for Radiotherapy Protectants and Chemotherapy
Protectants
26

27. Randomized Clinical Trials
Gold standard in Phase III
Single centre CT
– Primary and secondary indications
– Safety profile in patients
– Pharmacological / toxicological characteristics
Multi-centre CT
– Confirmation of the above
– Effect size
– Site, care and demographic differences
– Epidemiological determination
– Complexity
– Far superior to meta-analyzed determination of effect
27

28. Placebo Control Equality Trials
No anticancer drug treatment in the control arm is
unethical
Sometimes acceptable
– E.g., in early stage cancer when standard practice is to
give no treatment
– Add-on design (also for adjuvants)
all patients receive standard treatment plus either no
additional treatment or the experimental drug
– Placebos preferred to no-treatment controls because
they permit blinding
– Unless very low toxicity, blinding may not be feasible
because of a relatively high rate of recognizable
toxicities
28

29. Reasons for Active Control
1. Ethics – For trials involving mortality or serious
morbidity outcome, it is unethical to use placebo when
there are available active drugs on the market
2. Assay sensitivity – In trials involving psychotropic drugs,
placebo often has large effect. An active control is
sometime used to demonstrate that the trial has assay
sensitivity
3. Comparative purpose – To show how the experimental
drug compares to another known active drug or a
competitor
29

30. Non-Inferiority Trials
New drug not less effective by a predefined
amount, the noninferiority (NI) margin
– NI margin cannot be larger than the effect of
the control drug in the new study
– If the new drug is inferior by more than the NI
margin, it would have no effect at all
– NI margin is some fraction of (e.g., 50 percent)
of the control drug effect
30

31. Drug or Therapy Combinations
Use the add-on design
– Standard + Placebo
– Standard + Drug X
Effects seen in early phases of development
– Establish the contribution of a drug to a
standard regimen
– Particularly if the combination is more effective
than any of the individual components
31

32. What to Measure?
Primary outcome measure: The health parameter measured in all
study participants to detect a response to treatment. Conclusions
about the effectiveness of treatment should focus on this
measurement.
Secondary outcomes measure: Other parameters that are
measured in all study participants to help describe the effect of
treatment.
Baseline variables: The characteristics of each participant
measured at the time of random allocation.
– This information is documented to allow the trial results to
be generalised to the appropriate population/s
– Specific characteristics associated with the patient’s
response to treatment (such as age and sex) are known as
prognostic factors
32

33. What to Measure? E.g., Cancer Trials
Time to event end points
– Survival
– Disease free survival
– Progress (of disease) free survival
Objective response rates
– Complete
– Partial
– Stable disease
– Progressive disease
Symptom end points
Palliation
QoL
33

34. Cancer Trials – End Points
Endpoint Evidence Assessment Some Advantages Some Disadvantages
Survival Clinical benefit • RCT needed • Direct measure of • Requires larger and
• Blinding not benefit longer studies
essential • Easily measured • Potentially affected by
• Precisely crossover therapy
measured • Does not capture
symptom benefit
• Includes noncancer
deaths
Disease-Free Surrogate for • RCT needed • Considered to be • Not a validated survival
Survival (DFS) accelerated • Blinding clinical benefit by surrogate in most settings
approval or preferred some • Subject to assessment
regular • Needs fewer bias
approval* patients and shorter • Various definitions exist
studies than survival
34

35. Interim Analysis
35

36. Interim analysis
after each new response or group of responses
an interim analysis is performed
⇓
enough evidence to stop the trial
or
continue the trial
→ continuous sequential or group sequential analysis
36

37. Why interim analyses?
Ethics: superiority of a treatment
Safety: inferiority of a treatment /
toxicity of a treatment
Economy: = costly therapy
= no clinically relevant
difference in effect
between treatments
37

38. False Positives in Interim Analyses
Interim analysis for a trial in non-Hodgkins lymphoma; n=130,
IA after enrolment of each 25 patients
Response Rate Response Rate
CP CVP
Analysis 1 3/14 5/11 1.63
Analysis 2 11/27 13/24 0.92
Analysis 3 18/40 17/36 0.04
Analysis 4 18/54 24/48 3.25 0.05<P<0.1
Analysis 5 23/67 31/59 4.25 0.016<P<0.05
CP=Cytoxan-prednisone
CVP=Cytoxan-vincristine-prednisone
Source: Stuart J. Pocock, Clinical Trials
38

39. Interim Analysis of Data
2 Looks 3 Looks
How many times
0.05
0.05
can you look into
Nominal Pvalue
Nominal Pvalue
the data?
0.03
0.03
o pocock
o ob+fle
0.01
0.01
o fle+har+ob
0.0
0.0
1 2 1 2 3
Look Look
4 Looks 5 Looks
0.05
0.05
Type 1 error at kth
Nominal Pvalue
Nominal Pvalue
test is NOT the
0.03
0.03
same as the
0.01
0.01
nominal p value
0.0
0.0
for the kth test 1 2 3 4 1 2 3 4 5
Look Look
39

40. Considerations for IA
– Stopping rules for significant efficacy
– Stopping rules for futility
– Measures taken to minimize bias
– A procedure/method for preparation of data for analysis
– Data has to be centrally pooled, cleaned and locked
– Data analysis - blinded or unblinded?
– To whom the interim results will be submitted?
DSMB
Expert Steering Group
– What is the scope of recommendations from IA results?
– Safety? Efficacy? Both? Futility? Sample size
readjustment for borderline results?
40

41. 41

42. Equality Designs
(e.g., 2-Sample, Parallel)
H0 : Є = 0 HA : Є ≠ 0
Reject H0 when
p1— p2
ˆ ˆ
> z/2
√p1(1—p1)/n1 + p2(1— p2)/n2
ˆ ˆ ˆ ˆ
Where p1— p2 are true mean response rates
ˆ ˆ
from Test & Control
42

43. Superiority/Non-Inferiority Designs
(e.g., 2-Sample, Parallel)
H0 : Є ≤ δ HA : Є > δ … Superiority
H0 : Є ≥ δ HA : Є < δ … Non-Inferiority
Reject H0 when
p1— p2 – δ
ˆ ˆ
> z
√p1(1—p1)/n1 + p2(1— p2)/n2
ˆ ˆ ˆ ˆ
43

44. Survival Data – The Kaplan-Meier Estimator
1.0
0.75
Survival
0.50
0.25
~80% Patient
will survive
beyond 0.35
0.0
years Time (Year)
0.0 0.2 0.4 0.6 0.8 1.0
44

45. Include all Patients: ITT
It can be justified to look at data and drop the
“outliers”, poor compliers, inappropriately entered
patients
It is even plausible and acceptable as an academic
principle
But if not rigorously planned, such exclusion can
lead to bias
Even when planned, it can lead to imbalances that
also introduce bias
45

46. Intent-to-Treat Principle
All randomized patients
Exclusions on prespecified baseline criteria permissible
– also known as Modified Intent-to-Treat
Confusion regarding intent-to-treat population: define and agree upon
in advance based upon desired indication
Advantages:
– Comparison protected by randomization
Guards against bias when dropping out is realted to outcome
– Can be interpreted as comparison of two strategies
– Failure to take drug is informative
– Refects the way treatments will perform in population
Concerns:
– “Difference detecting ability”
46

47. Per Protocol Analyses
Focuses on the outcome data
Addresses what happens to patients who remain on therapy
Typically excludes patients with missing or problematic
data
Statistical concerns:
– Selection bias
– Bias difficult to assess
47

48. Intent to Treat & Per Protocol Analyses
Both types of analyses are important for approval
Results should be logically consistent
Design protocol and monitor trial to minimize exclusions
Substantial missing data and poor drug compliance
weaken trial’s ability to demonstrate efficacy
48

49. Missing Data
Protocol should specify preferred method for
dealing with missing primary endpoint
– ITT
e.g., treat missing as failures
e.g., assign outcome based on blinded case-by-case
review
– Per Protocol
e.g., exclusion of patients with missing endpoint
49

50. Multiple Analyses
The two main problems introduced by multiple analyses
are
– firstly, the increased probability of detecting
intervention effects where none exist (“false positives”
owing to multiple comparisons — type I errors)
– secondly, the limited capability (“power”) of trials to
detect a true treatment effect in secondary outcomes if
not enough participants are enrolled to show a
statistically significant difference in these outcomes
(“false negatives” — type II errors)
50

51. Assay Sensitivity
The critical question is whether a non-inferiority trial, for
example, could distinguish the control from placebo and shown
an effect of the non-inferiority margin
If it could
– the trial is said to have said to have “assay sensitivity”
If a trial a trial has assay sensitivity
– then if C-T < M, T had an effect
If the trial did not have assay sensitivity
– then even if C-T < M, we have learned nothing
If you don’t know whether the trial had assay sensitivity, finding no
difference between C and T means
– Both drugs were effective
– Neither drug was effective
Source: RJ Temple, US FDA, Unapproved Drugs Workshop
51

52. Assay Sensitivity: Major Problem in NI
In a non a non--inferiority trial, the trial itself does not show
the study’s ability to distinguish active from inactive therapy.
Assay ability to distinguish active from inactive therapy
Assay sensitivity must, therefore, be deduced or assumed,
based on
1. historical experience showing sensitivity to drug effects
2. a close evaluation of study quality and, particularly
important
3. the similarity of the current trial to trials that were able to
distinguish the active control drug from placebo
In many symptomatic conditions, such as depression, pain,
allergic rhinitis, IBS, angina, the assumption of assay
sensitivity cannot be made
Source: RJ Temple, US FDA, Unapproved Drugs Workshop January 2007
52

53. Data Safety and Monitoring Board (DSMB)
All trials may not need a DSMB
DSMB Membership
– Medical Oncologist, Biostatistician and Ethicist
Statistical expertise is a key constituent of a DSMB
Three Critical Issues
– Risk to participants
– Practicality of Periodic Review of a Trial
– Scientific Validity of the Trial
53

54. Conclusions
Randomized Clinical Trilas are very sophisticated and complex
Principal Investigators’, Trial Monitors’and Biostatisticians’ roles are
invaluable
Higher Phase (Phases 2, 3) Clinical Trials provide for the main
evidence of efficacy and safety
Clinical data is very complex (confounded, censored, skewed, often
fraught with missing data point), therefore, proper hypothesization
and statistical treatment of data are required
Prospective RCTs are usually the preferred approach for evaluation
of new therapies
54

55. Conclusions
Clinically meaningful margins must be well defined in Control trials
prospectively
– Superiority and non-inferiority margins must not be confused
Two or one-sidedness of α should also be prospectively defined
Power must be adequate
Variance must be analysed using the right model
Strategy for dealing with multiple end points must be prespecified
– Too many end points ot tests will increase the false positive (α) error
Sometimes (e.g., in equality trials) statistically significant results may not be
medically significant
Data censoring or skewed data must be well defined
– E.g., time to event data
55

56. Conclusions
Randomisation and blinding offer a robust way to remove bias in end-point
estimations
Data must be accurately captured without any bias and analysed by prospectively
described methods
Interim analysis should carefully plan ‘ spending’ function and the outcome
measure to be analyzed
Final analysis should be done carefully, independently and meaningfully (medical
as well as scientific)
Choose clinically relevant delta
Design, conduct, and monitor trials to minimize missing data and poor compliance to
drug
Analysis
– Both intent-to-treat and per protocol analyses should be conducted
– Sensitivity analyses
Outstanding medical and statistical issues must be brought to the fore.
56

57. Thank you
57

58. Order of the Topics
Clinical Trials
– Regulatory phases 1,2,3 & 4
Review of Clinical Trials
Data Requirements
Study Design Considerations
Controls
– Placebo, Active, NI
Assay Sensitivity, Multiplicity of Analyses
Interim Analysis
Intent to Treat Analysis
Final Analysis
Conclusions
58