Peginterferon ?2a/Ribavirin FDA Antiviral Drugs Advisory Committee

1
Peginterferon ?2a/Ribavirin FDA Antiviral Drugs
Advisory Committee

• Brian Murphy, MD, MPH, MS
• InterMune
• Submitted November 12, 2002 for
• Presentation November 14, 2002

2
Disclosure

• Presently serve as Vice-President Corporate
  Development/Hepatology at InterMune
• InterMune develops drugs in the HCV therapeutic
  area including interferons

3
What Information is Still Pending for the US
Treating Community?

• Absolute number of US patients in the
  peginterferon ?2a/ribavirin registration trial
• The withdrawal rate of US versus Ex-US patients
  in this registration trial
• Percentage response of US versus Ex-US patients
  enrolled in this registration trial
• Safety profile of US versus Ex-US patients
  enrolled in this registration trial

4
What Data Does the Treating Community Have Access
to?

• NEJM Paper 56
• On-therapy analysis
• Used data from weeks 68-72
• EU Analysis 54 / 50
• 54 was an on-therapy analysis
• Used data from week 60
• 50 was an intent-to-treat analysis
• Borderline statistical significance versus the
  comparator group p 0.057)
• Analyzed data per protocol

Fried MW, et al. NEJM, 2002 347(13)
975-982 European Union Scientific Discussion
Pegasys EMEA 2002
5
Impact of Patient Withdrawals Absolute number
of US patients
NEJM EU DISCUSSION
Total number of patients 453 468
Number who completed follow-up 334 337 (computed)
Number withdrawn for insufficient response 34 35 (computed)
Number withdrawn for other reasons () 85 (19 reported) 96 (computed)
6
Clinical Questions Absolute number of patients

• Did the withdrawal rate exceed that planned for
  the study?
• Most studies plan for a 10-15 drop-out rate
• Sample size power calculations are missing from
  the NEJM article were detailed in the
  peginterferon ?2b/RBV study article
• If the withdrawal rate exceeded the planned
  drop-out rate, are the results, including the
  safety data, significant?
• What was the absolute number of American patients
  that withdrew from the study?
• Is the baseline American presence more diminished
  by the withdrawal of US patients?

7
Treatment Response by Geographic Location
Peginterferon alfa-2b/RBV
Study Center Peginterferon Alfa-2b 1.5/RBV Intron A/RBV
US 169/346 (49) 131/337 (39)
Non-US 95/165 (57) 100/168 (59)
FDA Advisory Panel on Peginterferon alfa-2b
8
Prevalence of HCV Infection,United States,
1988-1994
African-American
Mexican American
Anti-HCV positive ()
White
12-19
20-29
30-39
40-49
50-59
6-11
60-69
70
Age group (years)
Adapted from Centers for Disease Control and
Prevention (M. Alter). MMWR. 199847(RR-19)4.
9
Prevalence () of Overweight/Obesity Among Males
in US, 1999-2000
Overweight BMI ? 25 Obesity BMI ? 30 Caucasians African-American Hispanic
Ages 20-39 61/22 52.6/27.4 67.5/30.4
Ages 40-59 69.9/28.5 63.9/29.9 79.1/27
Ages ? 60 74.3/34.3 69.1/26.4 79.6/29.7
Flegal KM, et al. JAMA 2002 288(14) 1723-1727
10
If the numbers are available, how should they be
analyzed?

• Intent-to-treat (ITT)
• Includes all patients randomized
• Avoids comparing non-randomized cohorts
• If similar centers, patient populations, and
  inclusion/exclusion criteria, can help
  health-care providers and their patients compare
  across studies if head-to-head trials are lacking
  
• Provides estimates of treatment effects that are
  more likely to mirror those observed in
  subsequent practice

Guidance for Industry, (E9) Statistical
Principles for Clinical Trials Center for
Biologics Evaluation and Research, Sept. 1998, ICH
11
Roche Statement on Clinical Trial Analysis

• Intent-to-treat versus on-treatment analyses
  ITT analysis includes all patients assigned to
  each study group regardless of whether they
  dropped out of the study or switched therapies.
  On-treatment analysis includes only those
  patients who completed the study within their
  originally assigned groups. Therefore,
  on-treatment analyses fail to account for
  dropouts and switches, and their treatment
  success rates tend to be deceptively higher than
  those from similar ITT analyses

Evaluating Clinical Studies-What Clinicians
Need to Know Roche Labs 2001
12
Conclusions (1)

• Global trial results may not be reflective of the
  American experience. For informed treatment
  decisions, the following American data would be
  valuable to know
• The absolute number of US patients in the study
• US vs. Ex-US response rates by ITT analysis
• US vs Ex-US safety parameters by ITT analysis
• US vs Ex-US withdrawal rates by ITT analysis

13
Conclusions (2)

• To adequately compare data, we are in agreement
  with the ICH Guidelines that support the use of
  ITT analyses, and agree with the Roche position
  that data analyzed by other methods may lead to
  deceptively higher results.