Clinical Trial Design for MildModerate CAP: A Drug Development Perspective

1
Clinical Trial Design for Mild/Moderate CAP A
Drug Development Perspective

• Anti-infective Drugs Advisory Committee
• April 2, 2008
• Roger M. Echols, M.D.
• Chief Medical Officer
• Replidyne, Inc.
• Previous positions V.P. Clinical RD
  Infectious Diseases, Bristol Myers Squibb 1997
  2004
• Medical Director Anti-Infectives,
  Miles/Bayer 1989 - 1996

2
Points to Consider

• CAP is a continuum of disease from mild to severe
• Fine Class does not define infection severity,
  its a predictor of 30 day mortality
• S. pneumoniae is a leading cause of CAP for all
  Fine Classes
• Untreated CAP does not get better quickly
• Non-inferiority margins can be determined using
  clinical response, not just mortality
• Historical data
• Contemporary trials showing superiority
• If treatment benefit (M1) is large, what matters
  most is M2 which is based on clinical judgment

3
Superiority Trials in CAP not Feasible

• 1992 FDAs Points to Consider
• Placebo control ethically unacceptable not to
  treat infected patients when therapy is
  available
• Active control high cure rates make it nearly
  impossible or impractical for a new antimicrobial
  drug product to demonstrate statistical or
  clinically relevant superiority to an approved
  comparator agent
• Recent experience with placebo controlled studies
  in AECB and ABS - reasons for rejections
• placebo (even with rescue therapy) conflicts with
  treatment guidelines
• Without comparator, study of no value (unethical)

4
Clinical Trial Experience in CAP 2000-2002

• Dataset of 7 CAP trials conducted globally
  N2128 subjects
• 6 NI studies 2 oral Rx only 2 IV/PO 2 optional
  IV/PO or oral
• 1 uncontrolled CAP oral Rx only
• All subjects characterized by Fine Class
• 2 oral Rx only studies restricted to Fine Class
  1-2 no restrictions for other studies
• Microbiology included 1257 typical pathogens 810
  atypical pathogens
• Fine Class distribution
• Class 1 46
• Class 2 32
• Class 3 13
• Class 4 8
• Class 5 1.5

78
5
Distribution of Respiratory Pathogens by Fine
Class Clinical Trial Experience
a-percentages are based on the total number of
pathogens in each Fine Class Category B
Nnumber of subjects in each Fine Class
6
Large Mortality Benefit of Chemotherapy in
Pneumonia (CAP) - Historical Data

• Evans, G.M. and Gaisford W.F. - 1938 Lancet
• 200 cases of lobar pneumonia admitted March -
  June 1938
• Not randomized, treatment allocation based on
  admission ward
• Variable dosing with 2-(p-aminobenzenesulphonamide
  ) pyridine (sulfapyridine)
• Case mortality rate - treated (n100) - 8
• untreated (n100) - 27

Evans, G.M. and Gaisford W.F. - 1938 Lancet,
pages 14-19
7
Rapid Clinical Response to Chemotherapy in CAP -
Historical Data, Flippin, et al.

• 100 cases sulfapyridine only (25 grams total)
• All serotyped S.pneumoniae (8 with bacteremia)
• Case mortality 4 (3/4 Type III)
• Temperature response
• 24 hours 46
• 48 hours 37
• 120 hours all patients
• Summary a conspicuous effect of the drug seemed
  to be its ability to bring about, within
  twenty-four to forty-eight hours, a critical drop
  in temperature followed by prompt clinical
  improvement

83 lt 48 hours
Flippin, H.F. et al. JAMA, 1939 112529-34
8
Untreated Pneumococcal Pneumonia Response by
Crisis Cecil, R, Textbook of Medicine, 1942
Temp Pulse
9
Pneumococcal Pneumonia Response to
Sulfapyridine Cecil, R, Textbook of Medicine,
1942
10
Historical Data Clinical Response

• Bullowa dataset from Management of the
  Pneumonias, 1937
• Cohort of 662 untreated patients with S.
  pneumoniae pneumonia who spontaneously recovered
  (survived)
• Day of initial response or crisis recorded
  mode day 8-9, only 3 by day 3 14gt2 weeks
• Agranat study from The Lancet, 1938
• Four hospitals in S. Africa 550 patients
• Treatment allocated by admission ward
• Day of pyrexia termination recorded at day 3,
  70 vs. 15 responded

11
Initial clinical response in untreated patients
(N662) with confirmed pneumococcal pneumonia
of patients
Adapted from Bullowa, 1937
days
12
Large Treatment Benefit in Clinical Response
Observed with Sulfapyridine in 1938
Time to pyrexia termination among sulfapyridine
treated or untreated patients (N550) with
community acquired pneumonia, South Africa 1938
of patients
Difference _at_ day 3 55.6 with 95 CI (48.5,
62.7)

Adapted from Agranat et al., 1938
days
13
Measuring Clinical Response in CAP

• Post treatment physician assessment
• global response, takes into account early relapse
• established dataset for comparison
• Patient Reported Outcome (PRO)
• Lamping et al is not a PRO, its a patient
  questionnaire
• Not sensitive to differentiate between active
  treatments
• No historical data upon which to base a NI margin

14
CAP Symptom Questionnaire Score Does Not
Discriminate CAP 2000 Study
Score
Moxifloxacin
Standard treatment
Source Table 5 in Effectiveness of oral
moxifloxacin in standard first-line therapy in
community-acquired pneumonia, Eur Respir J 2003
21135-143 (Torres et al.)
15
A New Primary Endpoint in CAP

• Composite clinical response that captures
• Early response _at_ 72 hours
• Objective measures, i.e. temperature, WBC,
  respiratory rate, cough
• Confirmed by TOC clinical response
• Mortality failure
• Represents physician assessment with objective
  criteria
• Minimizes individual investigator bias
• Includes early clinical response
• Captures historical evidence of treatment benefit

16
Superiority Trial in Mild/Moderate CAP
Levofloxacin vs. Ceftriaxone/cefuroxime

• Multicenter, prospective, randomized, open label
• Levofloxacin 500 mg q.d. (IV or PO) x 7-14 days
• Ceftriaxone (1 or 2 gram) q.d. or cefuroxime 500
  mg bid x 7-14 days
• Patients ambulatory or hospitalized
• Diagnosis sputum and blood culture, DFA and
  urinary antigen for L. pneumophila, serology for
  atypicals
• 590 patients (295/arm) 53 outpatient 84
  mildmoderate infection
• Levo 61 oral only cef/cef 50.4 oral only
• Population Levofloxacin Cef/Cef Diff 95 CI
• Clinical Eval1 95 83 12 (6.2, 18.6)
• Micro Eval1 96 80 16 (7.4, 23.5)
• 1. FDA. Medical Officers review, NDA 20-634
  Levaquin Tablets. 1996

17
Defining the NI Margin M2 and Clinical Judgment

• Evidence exists for large treatment benefit for
  clinical response
• M1gt 50 for early clinical improvement
• If large treatment benefit, M2 should be based on
  clinical judgment how much less effective than
  standard of care is clinically acceptable?
• A new drug that is not much worse than standard
  of care may have advantages
• Treatment of resistant pathogens
• Better safety or tolerability
• Improved compliance
• Decrease pressure for resistance to current drugs
• All about benefit-risk, i.e. clinical judgment

18
What Population for Analysis?Impact on Sample
Size

• FDA prefers co-primary analysis for NI trials
• Previously Clinically evaluable (CE) and ITT
• Currently CE and mITT (ITT with positive micro)
• Clinically evaluable population 85 of enrolled
  subjects
• mITT 30-35 for typical pathogens
• Study powered for CE population _at_10? N484
• Study powered for mITT population _at_15?
  N556 (if 10? N1172)
• Two studies required for approval of CAP

19
Mild/Moderate CAP An Industry Perspective

• CAP is a continuum of disease
• Superiority trials in mild/moderate CAP
• placebo not possible
• active control high risk, unethical if use
  substandard Rx
• Non-inferiority margin can be determined using
  clinical judgment and statistical reasoning
• Impact on drug development
• population for analysis impacts sample size
• CAP only one indication, but anchor for RTI
  indications
• Study design clarity and acceptable feasibility
  is critical. Without clarity investment in new
  antimicrobials will diminish further