Basic Requirement of Law and Regulations

1
Basic Requirement of Law and Regulations

• Robert J. Temple, M.D.
• Associate Director for Medical Policy
• Center for Drug Evaluation and Research
• U.S. Food and Drug Administration

UNC School of Public Health March 6, 2008
2
Law, Regulations, Guidance

• Law and History
• Critical Regulations
• Critical Guidance

3
The Ages of Drug Development and Drug Regulation

• Age of Safety - 1938
• Age of Effectiveness - 1962
• Age of Individualization and Dose-Response - from
  early 1980s
• AGES CUMULATE AND CONTINUE TO DEVELOP AS WE LEARN

4
The Ages of Drug Development and Drug Regulations

• Pre-History (pre-1938)
• FDA existed (1906) but could only respond to
  problems. There was complete freedom to market
  no requirement for testing or approval
  Government could seek to remove dangerous or
  misbranded products. There were some disasters
• DNP - weight loss drug, caused thousands of
  cataracts, enucleations in 1930s
• 1937 - Elixir sulfanilamide killed over 100
  children diethylene glycol (anti-freezes) - no
  animal tests, led to
• The Age of Safety the Federal Food, Drug and
  Cosmetic Act of 1938

5
The Food, Drug and Cosmetic Act of 1938

• Required
• 1. Pre-market notification. Marketing required
  an approved new drug application (NDA), but the
  NDA became effective if FDA did not object time
  could be extended. Reflected a strong
  expectation that there would be approval

6
The Food, Drug and Cosmetic Act of 1938

• 2. Required a demonstration of safety. The
• application could be refused if
• (a) Investigations did not include all tests
  reasonably applicable to show whether drug is
  safe when used under proposed labeling
• (b) Results of tests show unsafe or do not show
  that it is safe
• (c) Information submitted or any other
  information available are
• insufficient to determine whether safe
• The safety requirements of 1938 are identical to
  current requirements. Note how broad and
  possibly subject they are

7
The Food, Drug and Cosmetic Act of 1938

• (d) Labeling is false or misleading in any
  particular
• The safety requirements of 1938 are identical to
  current requirements. Note how broad and
  possibly subjective they are

8
The Age of Effectiveness

• People learning and practicing medicine today
  cannot really imagine what the basis of medicine
  was like even in the 1950s and 1960s. Apart
  from obvious things (removing an appendix, curing
  infections, Lasix), it was hard to say what we
  really knew
• Controlled trials were hardly ever seen
  effectiveness was rarely convincingly established
• Drug labeling was fantasy (Early 1960s PDR
  listed 50 treatments for alcoholism)
• Outcome trials basically didnt exist till VA
  studies in HT (1967), UGDP (late 1950s)
• Statistical inference was primitive once you left
  a few special places (NIH, mostly, and perhaps
  analgesia)
• And then it all changed, beginning in 1962

9
The Age of Effectiveness - 1962

• Why thalidomide (a safety problem) led to a
  change in how to recognize effectiveness is not
  obvious, but it put the FDC Act in play and
  then anything can happen. Actually, the 1962 Act
  made at least 3 important changes
• 1. FDA had to give positive approval before a
  drug could be marketed
• 2. A meaningful requirement to study drugs under
  an IND and explicit requirement for informed
  consent
• 3. The effectiveness requirement
• It also required review of all the drugs approved
  1938-62 to determine effectiveness. A huge, but
  successful, effort started by a contract with the
  NAS/NRC, leading to withdrawal of about one third
  of existing drugs and many claims for the ones
  that remained

10
The Effectiveness Requirement

• An NDA can be rejected if
• There is a lack of substantial evidence that the
  drug will have the effect it purports or is
  represented to have under proposed labeled
  conditions of use (this is what an applicant must
  show)
• The Law then goes on to describe what substantial
  evidence is. It is evidence consisting of
  adequate and well-controlled investigations,
  including clinical investigationson the basis of
  which it could be concluded that the drug will
  have the effect it is represented to have under
  the conditions of use proposed in labeling (this
  is how the applicant must show effectiveness)

11
The Effectiveness Requirement

• It was the only new requirement for approval in
  1962
• It was not the effectiveness requirement that was
  radical (safe for intended use could alone imply
  a risk/benefit analysis, i.e., need for evidence
  of benefit) it was the need for adequate and
  well-controlled studies that changed everything,
  all of medical science, really
• These are the only basis for approval
• Note the plural. Agency, supported by
  legislative history,
• interpreted this as requiring more than
  one controlled trial
• (modified by FDAMA 1997 to allow one study
  in some cases)
• No relative efficacy (unless inferior
  effectiveness leads to lack of
• safety)
• Effect must be clinically meaningful (added by
  Court)
• Must provide all relevant information. NOT
  SUMMARIES

12
The Effectiveness Requirement (cont.)

• It was really an amazing stroke
• In those days (not any more), laws tended to be
  general, leaving details to the agencies with
  expertise. That philosophy would lead to the
  substantial evidence requirement, not further
  defined
• For Congress to go further and say what the only
  kind of acceptable study could be was remarkable
• Actually a very clever trade-off. Substantial,
  legally, is a low standard (between a scintilla
  and a preponderance)
• But adding a need for two AWC studies turns a
  low standard into quite a high one especially
  with the p lt 0.05 (two-sided) that emerged

13
Labeling

• Labeling must bear adequate directions for use
  and may not be false or misleading
• Very critical to support requirements for
  dose-response and individualization information
  (although they also relate to safety and
  effectiveness)
• That is the totality of the legally mandated
  clinical requirements. The rest is regulation
  and, even more, guidance and practice

14
The Effectiveness Requirement (cont.)

• In 1962, of course, and really until 1970 or so,
  we at FDA had only a poor idea of what a
  well-controlled study was, and things we take for
  granted now were not at all known. But we
  learned for example, about
• 1. Interim looks at data, multiplicity
• 2. Counting all patients, cause-specific
  mortality
• 3. Interpretation of active control
  non-inferiority trials
• 4. Good dose response

15
Counting All Patients

• In 1980 the need to account for all patients in a
  trial had never been an issue. Then came the
  Anturane Reinfarction Trial, a study, supported
  in the NEJM by two Dr. Braunwald editorials, that
  seemed to show a survival benefit in post-AMI
  patients treated with sulfinpyrazone (Anturane),
  an anti-platelet drug. Our analysis taught us a
  lot about cause-specific mortality, multiple
  endpoints, unplanned 6 month subset analyses and
  complete follow-up Temple R, Pledger G. The
  FDA's Critique of the Anturane Reinfarction
  Trial. N Engl J Med 3031488-1492, 1980

16
Anturane Reinfarction Trial (ART)

• Randomized comparison of sulfinpyrazone
  (Anturane) and placebo in 1500 post infarction
  (25-35 days) patients.
• Distinguished trialists
• An ambitious industry-funded trial

17
(No Transcript)
18
(No Transcript)
19
Ineligible Patients

• It was not possible to see this from published
  reports, but 9 patients who had died were
  excluded from the results (8 Anturane, one
  placebo) for being ineligible or poor
  compliance (pills found in their room). When you
  put back exclusions, there was no documented
  effect.

20
(No Transcript)
21
Counting All Patients (cont.)

• FDA guidance now clear in calling for full
  patient accounting. CONSORT statement similar.
  There is no doubt the 6 missing ineligible deaths
  had not been reported in NEJM
• Weve also learned to be careful about
  cause-specific CV mortality. Cause specific
  mortality concern triggered by change in one
  patient cause from interim to final, triggering a
  You mean these can change thought and review of
  cases. We found that similar events were
  reported as SD on placebo, something else on
  Anturane, leading to a reduced SD claim. Here
  too we have been very cautious about such
  analyses
• Finally, transformation of an all CV death
  analysis to a sudden death in 6 months would be
  greeted skeptically

22
Regulations

• The clinical parts of an application are affected
  mainly by 3 regulations, the first 2 revised in
  1985, the third created in 1992
• 21 CFR 314.50 Content and Format of an
  application
• 21 CFR 314.126 Adequate and Well-controlled
  studies
• 21 CFR 314.500 Accelerated approval (use of
  surrogate endpoints and approval with
  restrictions)

23
Content and Format of an Application (21 CFR
314.50)

• A. Summary, including annotated labeling. There
  is existing guidance but the new ICH Common
  Technical Document format (CTD) will replace it
• B. Technical sections also substantially
  altered by CTD
• 1. Chemistry
• 2. Non clinical pharm/tox
• 3. Human PK and bioavailability
• 4. Microbiology
• 5. Clinical
• 6. Statistical section

24
Clinical Section

• Rule calls for
• 1. Description and analysis of every clinical
  pharmacology study and every controlled study,
  including the protocol and statistical analysis,
  as well as sufficient reports of everything else
  that is pertinent to safety and effectiveness
  from any source

25
Clinical Section

• 2. Integrated summary of data showing substantial
  evidence of effectiveness and evidence to support
  dosage and administration, modifications for
  subgroups (pediatrics, geriatric, renal failure)
• 3. Summary and updates (4 months prior to
  approval) of safety information with all
  available information related to safety,
  including animal data, adverse effects, drug-drug
  interactions

26
Clinical Section

• 4. Case report forms for each patient who died or
  did not complete study because of an adverse
  event (thought drug related or not). Others on
  request. Prior to 1985, all CRFs required
• 5. Case report tabulations (replaced all CRFs)
• All data from well-controlled studies
• All data from earliest clinical pharmacology
  studies
• Safety data from other studies
• Original intent was archival moving toward
  usable data sets

27
Adequate and Well-Controlled Studies

• 314.126 Adequate and Well-Controlled Studies
• Very critical, describes the critical features of
  the only kind of studies that can support
  approval.
• Apart from design and analysis (A and WC) must
  show effectiveness, ordinarily a statistically
  significant effect on a meaningful endpoint,
  usually replicated.

28
Adequate and Well-Controlled Studies

• Directed at three main goals
• 1. Need a valid control group because the course
  of a disease is variable the state of the
  disease can change spontaneously, and is subject
  to many influences. The control group, a group
  very similar to the test group, and treated the
  same as people getting the test drug, except for
  getting the drug, lets you tell drug effect from
  other influences, such as spontaneous change,
  placebo effect, biased observation.
• (If course was predictable, you would just
  intervene and observe.)

29
Adequate and Well-Controlled Studies

• 2. Need to minimize bias, a tilt favoring one
  treatment group, a directed (non-random)
  difference in how test and control group are
  selected, treated, observed or analyzed
• 3. Sufficient detail to know how the study was
  done and what results were

30
Adequate and Well-Controlled Studies (Contd)

• Reports of adequate and well-controlled
  investigations provide the primary basis for
  determining whether there is substantial
  evidence to support the claims of effectiveness
  for new drugs and antibiotics. Therefore, the
  study report should provide sufficient details of
  study design, conduct, and analysis to allow
  critical evaluation and a determination of
  whether the characteristics of an adequate and
  well-controlled study are present.

31
Kinds of Controls

• Placebo control
• No treatment concurrent control
• Dose-response control
• Active Control
• Historical Control
• There is no hierarchy all types can be, and in
  any given year are, used as the basis for
  approval of a drug. But not every design is
  usable in every situation.
• Critical distinction trials that must show a
  difference to succeed (stimulus to excellence
  because sloppiness gives a bias toward the
  null) and trials intended not to show a
  difference (active control non-inferiority
  studies).

32
Dose-Response

• D/R study one kind of controlled trial positive
  D/R slope shows effectiveness.
• Also, growing recognition that it is important to
  choose a reasonable dose - ICH guideline 1993.
  Show D/R for benefits and risks.
• Historical error diuretics
• Effective dose 1/8-1/4 dose used
• Hypokalemia, probably decreased benefit of
  treatment
• Disparity between stroke effect (40) and cardiac
  effect (15
• until low-dose used (SHEP)

33
Dose-Response Studies

• Until early 1980s, most trials with more than
  one dose titrated the dose, generally to some
  endpoint. This meant
• 1. The group on any given dose was not chosen
  randomly
• 2. Time and dose were confounded secular trend
  would look like response to dose. Particularly
  useless for safety
• In 1980s, FDA promoted the randomized, parallel,
  fixed dose, dose-response study, identified as
  the standard in ICH E4 guidance. Note, D/R
  studies can serve two purposes
• 1. Show effectiveness
• 2. Show D/R

34
Adequate and Well-Controlled Studies (Contd)

• (IV) Active Treatment Concurrent Control. The
  test drug is compared with known effective
  therapy for example, where the condition treated
  is such that administration of placebo or no
  treatment would be contrary to the interest of
  the patient. An active treatment study may
  include additional treatment groups, however,
  such as a placebo control or a dose-comparison
  control. Active treatment trials usually include
  randomization and blinding of patients or
  investigators, or both. If the intent of the
  trial is to show similarity of the test and
  control drugs, the report of the study should
  assess the ability of the study to have detected
  a difference between treatments. Similarity of
  test drug and active control can mean either that
  both drugs were effective or that neither was
  effective. The analysis of the study should
  explain why the drugs should be considered
  effective in the study, for example, by reference
  to results in previous placebo-controlled studies
  of the active control drug.

35
Equivalence Trials

• A major regulatory, ethical, international
  problem
• Fundamental distinction between trials intended
  to show a difference and trials intended to show
  similarity latter pose major problems of
  interpretation.
• Desire to use equivalence is understandable
  seems sensible to compare new and old effective
  therapy, see no difference and declare victory.
  Avoids exposure to ineffective treatment.
• I will return to this in more detail later.

36
Adequate and Well-Controlled Studies (Contd)

• (V) Historical Control. The results of treatment
  with the test drug are compared with experience
  historically derived from the adequately
  documented natural history of the disease or
  condition, or from the results of active
  treatment, in comparable patients or populations.
  Because historical control populations usually
  cannot be as well assessed with respect to
  pertinent variables as can concurrent control
  populations, historical control designs are
  usually reserved for special circumstances.
  Examples include studies of diseases with high
  and predictable mortality (for example, certain
  malignancies) and studies in which the effect of
  the drug is self-evident (general anesthetics,
  drug metabolism).

37
Historical Control (External)

• Retrospective
• Unblinded
• Selection bias
• Past experience, other non-random experience
• Baseline (patient as own) control

38
Historical Controls

• Critical Reference -
• Sacks, Chalmers, Smith
• Am J. Medicine (1982) 72233-240.
• Comparison of RCTs and HCTs for same disease
• Always
• 1. RCT less favorable than HCT
• 2. Reason was that the historical control was
  worse than the randomized control (selection
  bias)
• 3. Not possible to adjust the difference
• Many examples of misleading HCTs great care in
  relying on one

39
Adequate and Well-Controlled Studies (Contd)

• (4) The method of assigning patients to treatment
  and control groups minimizes bias and is intended
  to assure comparability of the groups with
  respect to pertinent variables such as age, sex,
  severity of disease, duration of disease, and use
  of drugs or therapy other than the test drug.
  The protocol for the study and the report of its
  results should describe how subjects were
  assigned to groups. Ordinarily, in a
  concurrently controlled study, assignment is by
  randomization, with or without stratification.
• Bias reduction before the trial.

40
Adequate and Well-Controlled Studies (Contd)

• (5) Adequate measures are taken to minimize bias
  on the part of the subjects, observers, and
  analysts of the data. The protocol and report of
  the study should describe the procedures used to
  accomplish this, such as blinding.
• Bias reduction during and after the trial

41
Endpoints of Trials

• The choice of study endpoints is critical to drug
  assessment, but law and regulations say little.
  The endpoint must be clinically meaningful
  (Court) but can be
• important outcome death, AMI
• symptom
• surrogate endpoint
• A surrogate endpoint, or marker, is a
  laboratory measurement or physical sign that is
  used in therapeutic trials as a substitute for a
  clinically meaningful endpoint that is a direct
  measure of how a patient feels, functions, or
  survives and that is expected to predict the
  effect of the therapy

42
Accelerated Approval (21 CFR 314.500)

• Nothing in law forbids use of a surrogate
  endpoint for approval and some are considered
  valid and regularly use (BP, BS, cholesterol)
• But experience with antiarrhythmics, inotropic
  drugs for heart failure, and even the relatively
  modest effect of antihypertensives on CV
  mortality has led to considerable skepticism
• A rule (1992) on Accelerated Approval addressed
  this, reflecting both skepticism and the sense of
  urgency that can arise in relation to serious,
  untreatable illnesses Incorporated into FDAMA,
  1997

43
Accelerated Approval

• Approval based on a surrogate endpoint that is
  reasonably likely, based on epidemiologic,
  therapeutic, pathophysiologic, or other evidence
  to predict clinical benefit.
• Conditions
• 1. Serious or life-threatening illness
• 2. Meaningful therapeutic benefit over existing
  treatments
• 3. Requirement to study the drug post-approval to
  verify and describe its clinical benefit.
• 4. Easy removal
• Used principally for AIDS drugs (viral load, T4
  lymphocytes) and oncologic drugs (response rate
  in refractory disease)

44
II. Quantity of Evidence Needed to Support
Effectiveness

• A. Legal standard it has been FDAs position,
  based on the words of the statute (adequate and
  well-controlled studies) and legislative
  history, that Congress intended to require at
  least two adequate and well-controlled studies,
  each convincing on its own, to establish
  effectiveness. We have been supported by the
  Courts.
• But we have been flexible
• Relied on other pertinent studies (different
  stages, dosage forms, regimens) to support a
  single A WC study. (There are A WC studies
  in this case.)
• Relied on single excellent multicenter study with
  statistically strong finding, generally where
  there was an important clinical benefit, making a
  confirmatory study difficult to conduct on
  ethical grounds

45

• B. Scientific Basis for Legal Standard
• Independent Substantiation needed because
• A trial may have undetected biases
• Chance can yield a single favorable result. This
  can be dealt with by statistical evaluation but
  note the extent of multiplicity when there are
  hundreds of clinical trials
• Results may be site-dependent, not generalizable
• Fraud can occur
• Not necessarily (or even best) an exact
  replication we use the term independent
  substantiation, not replication

46
505(d), as amended

• ...As used in this subsection, the term
  substantial evidence means evidence consisting
  of adequate and well-controlled investigations,
  including clinical investigations...that lead to
  the conclusion that the drug will have the
  effect... it is represented to have... NEW. If
  the Secretary determines, based on relevant
  science, that data from one adequate and
  well-controlled clinical investigation and
  confirmatory evidence (obtained prior to or after
  such investigation) are sufficient to establish
  effectiveness, the Secretary may consider such
  data and evidence to constitute substantial
  evidence for purposes of the preceding sentence.

47
How Many Studies?orWhen Can an Effectiveness
Conclusion be Based on a Single Study

• Guidance Providing Clinical Evidence of
  Effectiveness for Human Drug and Biological
  Products (May 1998)
• Response to FDAMA (1997), though had been under
  development for several years

48
Guidance/Advice

• Law and regulations give only the most general
  description of what is needed. Details come in
  guidance, which comes in several forms
• 1. Guidance documents
• Specific clinical guidances for a therapeutic
  area
• More general guidance on approaches

49
Guidance/Advice

• 2. Meetings with FDA end of phase 2, other
• FDA always prepared to meet to help design
  clinical trials, agree on endpoints, design,
  number of studies, design, number of studies,
  size of safety data base, etc.
• Minutes record agreements, which we live by
  unless public health demands otherwise
• 3. Open Advisory Committee meetings
• 4. Availability of all reviews after drug is
  approved

50
Critical Guidance

• 1. Guideline on the Format and Content of the
  Clinical and Statistical Sections of New Drug
  Applications (1988)
• Includes guidance on reporting an important
  trial replaced by ICH E3 Structure and Content
  of Clinical Study Reports and on integrated
  analyses of the efficacy and safety data, which
  are required under 314.50

51
Critical Guidance

• 2. How to prepare the adverse reaction, clinical
  trials Warnings and Precautions section of
  labeling
• 3. Many Clinical Pharmacology Guidances,
  including
• Studies of drug-drug interactions, both in vitro
  and in vivo
• Studies in renal and hepatic impairment
• Developing exposure-response information (soon)
• How to do population kinetic studies

52
Critical Guidance

• 4. Other ICH Guidances
• E4 - Dose-Response Information to Support Drug
  Registration
• E5 - Ethnic Factors (Really, what additional data
  should be requested if submitted data are
  extra-regional)
• E6 - GCPs
• E9 - Statistical Principles for Clinical Trials
• E10 Choice of Control Group
• E-14 Clinical Evaluation of QT/QTc
• 5. Internal guidance review template and safety
  review