Summary of Public Response and Regulatory Perspective

1
Summary of Public Response and Regulatory
Perspective

• Katherine Laessig, M.D.
• Division of Antiviral Drug Products
• January 11, 2001

2
Presentation Outline

• Defining terms
• Summary of responses re
• study components (pt. popn, study regimen,
  endpoints, duration
• review of study designs (historical-controlled,
  open-label vs. blinded, intensification,
  concentration-controlled/dose-response,
  factorial)
• elaboration on 3 possibly useful designs (add-on,
  two-part hybrid, modified factorial)
• Regulatory Conclusions

3
Defining Terms

• Heavily Treatment Experienced (H.T.E.) Therapy
• a new/recycled drug REGIMEN used to treat pts who
  have experienced therapeutic failure (efficacy or
  safety)
• unlikely that a SINGLE new drug will suffice as
  salvage therapy
• for regulatory purposes, the contribution of a
  new DRUG to the regimen is of interest
• for clinical management strategies, the REGIMEN
  is the entity of interest
• H.T.E. ptsprevious rx with gt 2 HAART regimens
  containing gt 1 agent from each class (NRTIs,
  NNRTIs, PIs)

4
Defining Terms

• Drug of last resort vs. broad use
• crucial distinction because impacts overall drug
  development plan
• has activity but would be restricted for use
  only in H.T.E. because of toxicity, route of
  administration or other reasons
• in contrast to a first-in-class or next-in-class
  that can be used for both early rx or for H.T.E.

5
Summary of Responses Patient Population

• Broad representation of H.T.E. pts including
• low CD4 (lt50)
• high HIV RNA (gt100,000 copies/ml)
• Also include patients whose prior regimens have
  failed due to
• PK
• tolerability
• adherence

6
Summary of Responses Patient Population

• Stratification historically based on
• HIV RNA
• CD4 count
• Well-powered, randomized trials will control for
  heterogeneity rendering extensive stratification
  unnecessary
• Examination of pt. subsets useful for exploratory
  analyses

7
Summary of Responses Study Regimen

• Agreement regarding use of
• resistance testing to construct background
  regimen
• allow expanded access agents
• include pharmacologic enhancers
• Number of background agents
• flexibility in number of background agents
• use of pK enhancers does not count in total
• megaHAART may decrease tolerability and
  adherence and increase overlapping toxicities,
  drug interactions, and number of dropouts

8
Summary of Responses Study Duration

• Traditional approval based on demonstration of
  durability of viral suppression
• 48 week data
• Accelerated approval
• based on 24 week data
• consider earlier assessment of antiviral effect
  for this popn with longer term safety
• When should the determination of an antiviral
  effect (for these patients) be made?

9
Summary of Responses Virologic Endpoints

• Proportion undetectable may not be feasible
• except for multiple investigational or highly
  potent agents
• Alternatives
• Mean change from baseline in HIV RNA
• Proportion with gt X log drop in HIV RNA
• AUCMB
• Other suggestions?

10
Summary of Responses Clinical Endpoints

• Clinical endpoint
• new CDC class C events ( 20 conditions)
• Suggestions for alternative clinical endpoints
• including fewer Class C Events e.g. CMV, MAC
• composite endpoint of efficacy and
  safety/toxicity
• DAVDP perspective
• better collection and adjudication of clinical
  endpoints regardless of the primary endpoint or
  the pt. population
• difficult to weight toxicity for composite
  endpoint, i.e. nausea vs. CMV
• Agency needs to examine efficacy and safety
  separately for risk-benefit assessment

11
Summary of Responses Study Designs

• Historical-controlled
• Open label vs. blinded
• Intensification
• Concentration-controlled/dose-response
• Factorial

12
Summary of Responses Historical-controlled

• General agreement
• historical results not obtained in equivalent
  population
• heterogeneity of the H.T.E. pts
• evolving standard of care
• incomplete data from historical cohorts
• Natural hx of H.T.E. pts on failing or currently
  available salvage therapy regimens is not
  disputed

13
Summary of Responses Historical-controlled

• DAVDP position
• when concurrent control is feasible, single arm
  trials not advocated
• use of concurrent observational cohort data
  possible

14
Summary of Responses Blinded vs. Open Label

• Agreement that blinding all drugs is difficult
• pill burden
• unavailability of placebo
• use of resistance testing to choose OBR
• Partial blinding (test/control) is sufficient
• Statistical considerations for open label studies
• blinding reduces bias
• patient/physician expectations when assignment is
  known
• differential dropout (switches, loss to
  follow-up)
• need to account for these in analysis

15
Summary of Responses Blinded vs. Open Label

• Method to assess potential for differential
  dropout
• monitor subsequent enrollment of patients who
  discontinue trial into Expanded Access programs

16
Summary of Responses Intensification

• Intensificationadding on new agent to
  preexisting regimen with incomplete viral
  suppression
• Concerns
• may promote resistance as essentially monotherapy
• may exhaust an option that could have been used
  later
• Potential Uses
• acceptable if resistance develops slowly
• duration of intensification should be minimal and
  include early escape option

17
Summary of Responses Concentration-controlled
and Dose-response

• Community feedback
• avoids suboptimal levels (conc. controlled)
• higher levels overcome resistant mutants
• Industry concerns
• real time reporting for dose adjustment is
  difficult
• high intra- and inter-subject variability
• patient adherence may impact results
• unclear which specific exposure measurement is
  best correlated with response (for
  investigational drugs)
• MTD (maximally tolerated dose) should be used in
  this population

18
Response from Public Concentration-controlled
and Dose-response

• Dose-response
• previous use for registration
• to discern a rx effect, need to study doses on
  the steep part of dose-response curve
• some participants may receive suboptimal doses
• Higher doses may be necessary to suppress
  resistant virus
• illustrated by the following graph

19
Dose Response vs. Strain Susceptibility
Dose 3
Dose 2
Dose 1
?
log drug X
20
Response from Public Concentration-controlled
and Dose-response

• Concentration-controlled
• no precedent for registration purposes for
  antiretroviral agents
• assay considerations (unapproved not widely
  available)

21
Summary of Responses Factorial

• Industry concerns
• drug interactions
• overlapping toxicities
• timely availability of drug supply
• ownership of IND and data
• Community and FDA in favor of this approach
• industry concerns valid but not insurmountable
• factorial design can be modified
• randomized trial that participants are more
  likely to complete
• expenses for one trial shared by two or more
  companies
• blinding and provision of placebos easier

22
Three Examples for Discussion

• Noncollaborative studies of single
  investigational agents
• Add-on to OBR
• Two-part hybrid
• Collaborative design for gt1 investigational agent
• Modified factorial

23
1 Add-on to OBR

• OBR placebo vs. OBR study drug
• two arm trial for one investigational agent
• randomization and blinding preferred
• less desirable than modified factorial
• risk to pts of receiving placebo is decreased by
  early escape option

24
2 Two-Part Hybrid Design
Assess Contribution
Assess Durability
All receive OBRX
days
weeks
25
2 Two-Part Hybrid Design
Prospective Phenotypic Cohorts P.S. -log
lt 4 fold -2.2 4-10 fold -1.2 gt 10
fold -0.5 gt20 fold 0.1
OBR Drug X
26
2 Two Part Hybrid Design

• Two part refers to
• initial randomization prospectively defined
  observational cohort
• direct assessment of antiviral effect during
  first 10 d then indirect dose response via
  correlation with baseline PT
• Assumes that lead-in period is brief enough that
• pts on OLD Drug X dont develop resistance
• pts continuing OLD regimens wont have adverse
  consequences due to not changing therapies
• However, lead-in period needs to be long enough
  to assess antiviral effect
• May provide supportive evidence in NDA package

27
3 Modified Factorial Design

• Four arm trial for 3 investigational agents
  (Drugs A, B, and C)
• OBR A B
• OBR A C
• OBR B C
• OBR A B C
• Assumption is that OBR or OBR single study drug
  alone is inferior
• N is 33 less than needed for 3 separate trials
• The same active arm is compared against 3 control
  arms

28
Regulatory Conclusions

• Focus of todays meeting is DRUG specific and not
  centered on regimens or management strategy
• Contribution of a drug to safety and efficacy in
  broad pt populations as well as H.T.E. needs to
  be determined
• Caveats about these trials
• need to know drug interactions up front
• dose selection is extremely important and may be
  different for H.T.E.
• baseline resistance testing is useful for
  construction of OBR and for outcome analysis

29
Regulatory Conclusions

• Additional points to consider
• multiple agents make determination of adverse
  event causality for drug toxicity difficult
• trials of shorter duration affect safety database
• resistance may develop to first-in-class agents
  and compromise later virologic response to
  next-in-class agents

30
Regulatory Conclusions

• Provide data for spectrum of patients,
  particularly H.T.E. for first-in-class or
  promising next-in-class
• Important to consider strength of NDA package
• one controlled, traditional study plus other
  well-designed studies in H.T.E. may be preferred
  over 2 identical studies in naïve and less rx
  experienced patients
• Study designs must take into account targeted
  use-- H.T.E. vs. all patients