AntiInfective Perspective

1
Anti-Infective Perspective

• Anti-Viral Advisory Committee Meeting
• July 25th, 2000
• Alexander Rakowsky, M.D.
• Medical Team Leader/DAIDP/ODE4/CDER

2
Proper Perspective

• Focus can be on new drug development versus
  changes in dosing/formulation/combinations with
  approved drugs
• Focus can be on systemic agents versus topical
  agents
• Looking at approved, systemic agents in this
  presentation

3
Documents/Guidances

• 1992 Anti-Infectives Points to Consider and also
  the IDSA/FDA guidances
• Recent re-writes of guidance for various
  indications
• Clinical Effectiveness Document
• Again, focus here is on approved drugs with
  changes in dosing/formulation/combinations which
  lead to a NON-bioequivalent state

4
Outline of this Talk

• Dont kill the messenger
• Brief primer on PK/PD parameters used in
  antibacterials (HFD-520 and HFD-590)
• Discussion of how these parameters have or could
  be used
• Example

5
Basic Divide

• CONCENTRATION-DEPENDENT
• AUCMIC
• Peak ConcMIC
• Fluoroquinolones, aminoglycosides

• TIME-DEPENDENT
• TimeMIC
• Beta-lactams, vancomycin

6
MIC?????

• Mean inhibitory concentration
• MBC is similar concept
• Based on standardized in vitro work using
  specified conditions, growth media,
  concentrations, nutrient additives for fastidious
  organisms, etc.
• Reproducibility

7
MIC?cont

• Difficulty is in interpretation of MIC (S/I/R)
• Based on achievable drug levels (ADME parameters)
• Clinical data is of importance
• Defined after discussions by committee (NCCLS) or
  review (FDA)
• Even then, occasional disagreement

8
Time-Dependent

• Major parameter is TimeMIC
• Based on animal studies (e.g., Craigs work with
  AOM)
• Confirmed by clinical trials

9
Time-Dependentcont

• Time dependent on ADME parameters
• Cmax achieved
• distribution of drug (e.g. serum versus tissue,
  protein binding, etc.)
• half-life of drug (metabolism, excretion)

10
Time-Dependentcont

• MIC
• pathogen dependent
• resistant strains
• inoculum effect, effect of pH on activity, etc.

11
Time-Dependentcont

• Animal and human studies
• TMIC in 40-60 range is PREDICTIVE of clinical
  success
• 100 correlation? NO
• Varies also among members of the same class
• Other variables which need to be accounted for
  but not as well defined Post-antimicrobial effect

12
Concentration Dependent

• Major parameters are PeakMIC and AUCMIC ratios
• Animal studies conducted
• Human studies done with more recent FQs
  (Drusanos work)

13
Concentration Dependentcont

• Still dependent on ADME parameters
• absorption
• distribution
• local levels/penetration
• local effects (e.g., pH)
• Clinical studies predictive, but again, not 100
  correlations.

14
Conclusions so far

• Variables are based on ADME ranges
• Work has shown good predictiveness but not 11
  correlation
• Variability in the MICs
• Most work done on beta-lactams and FQs

15
So What is the Role of PK/PD?

• Several recent meetings to discuss
• DAIDP Advisory Committee 7/98 and 10/98
• July 1998 FDA/Industry meeting
• March 1999 FDA/ISAP Workshop
• ISAP The International Society of Anti-Infective
  Pharmacology

16
Other Difficulties Raised

• Emphasis has been on effectiveness,not safety
• Most work done with single drug/bug
• Acute models used (chronic use/chronic illness
  not well studied)
• Moving targets (resistance development), esp.
  with chronic use, use over time

17
Is All Lost?

• Overall, positive impressions
• PK/PD for certain antimicrobial drug classes is
  well worked out
• Models used (both animal and human) are improving
  greatly
• Can be seen, in the proper context, as strong
  supportive evidence

18
Coming Full Circle

• Augmentin 71 NDAs (submitted in 1994/1995)
• Adults 500 mg tid to 875 mg bid of amoxicillin
  and 250 mg tid to 500 mg bid
• Pediatrics 40/mg/kg/day divided tid to 45
  mg/kg/day divided bid of amoxicillin
• In all formulations, clavulanic acid amount
  remained the same. Led to 1/3rd less daily
  clavulanate

19
Augmentin (cont)

• In all settings, AUC and half-life comparable
  between new and old dosing regimens
• Cmax higher by 50-80 in bid dosing regimens
• TMIC lower in bid regimens
• on average, bid regimens with 10 hours (out of
  24)
• on average, tid regimens with 11 hours (out of
  24)
• concerns also with 1/3rd lower beta-lactamase
  inhibitor activity

20
Augmentincont

• Post-antibiotic and post-beta-lactamase inhibitor
  effects were studied and proposed
• Animal studies showed comparable efficacy rates
  for the bid and tid dosing regimens
• Due to concerns with lower TMIC and clavulanic
  acid, clinical studies were asked for

21
Augmentincont

• One study per indication was conducted
  (historically, two would have been required)
• NDAs were approved based on
• in vitro microbiology and animal work
• PK/PD data from humans
• one adequate, well-controlled study per
  indication
• agreement to study as q12 not bid

22
Augmentincont

• Ultimately these NDAs were approved with
  approximately 50 less subjects enrolled then
  historically required
• See this as a good example of where/how PK/PD
  parameters can be used

23
Conclusions

• Certain parameters/drug classes well worked out
• Still issues with variability (with ADME
  parameters, MICs, local effects, etc.)
• Multiple meetings held where, at this time, PK/PD
  is seen as strong supportive evidence but that
  for reasons listed, should not be used in lieu of
  clinical evidence