The Case for Global Collaboration in Studies of Inhibitors in Hemophilia

1
The Case for Global Collaboration in Studies of
Inhibitors in Hemophilia

• Mike Soucie, PhD
• Division of Hereditary Blood Disorders
• National Center on Birth Defects and
  Developmental Disabilities
• Centers for Disease Control and Prevention
• Atlanta, GA, U.S.A.

2
Inhibitor Occurrence Rates

• Recent clinical trials of recombinant FVIII
  products in PUPs found a cumulative incidence of
  25-32 that eventually fell to about 12
• Several national database analyses of incidence
  in PTPs (gt150 ED) found a rate of 1-3 over 2 or
  more years
• Most inhibitors in PTPs are low titer (lt5 BU) in
  contrast to greater percent high titer in PUPs

Key, NS. Br J Haematol 2004 127379-391
3
Risk Factors

• Hemophilia gene defect intron 22 inversions,
  large gene deletions and stop codons higher risk
• Family history risk highest for twin, then
  brother, much less in extended family
• Race African origin up to two-fold risk
• Other polymorphisms and mutations in other
  immune response genes, inflammatory state,
  early exposure to factor

4
Inhibitor Research Challenges

• Characterization and risk assessment of rare
  adverse events is methodologically difficult
• Evaluation of treatment product risk using a
  probabilistic model is imperfect
• Clinical trial methodology is problematic
• Sample size not adequate to fully assess risk
• Incomplete assessment of risk factors
• Randomization to larger population is problematic

5
Universal Data Collection Program

• Purpose
• Monitor blood safety among recipients of blood
  products
• Monitor the extent and progression of
  joint disease
• Identify issues for further study

6
Participating HTCs
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Guam
?
?
?
?
?
?
?
Puerto Rico
7
Universal Data Collection (UDC)Study Design

• National protocol approved by human
  Investigational Review Boards. Participation is
  voluntary and patients give informed consent
• Standardized data collection annually using tools
  designed with input from bleeding disorder
  experts
• Blood specimen is tested centrally for known
  infectious disease agents (hepatitis and HIV)
• New infections are investigated for link with
  product
• Remainder of blood specimen is stored for future
  use

8
UDC Enrollment

• Since May, 1998, 11,023 people with hemophilia
  have been enrolled
• Represents 87 of the estimated 12,600 persons
  with hemophilia receiving care in HTCs in the
  United States
• Data have been collected from more than 25,000
  comprehensive care visits made by persons with
  hemophilia
• Participation rates have been greater than 90

9
UDC Data On Inhibitors

• Although UDC was not designed to study
  inhibitors, some related data are collected
• Demographic information
• Bleeding frequency (measure of exposure)
• Inhibitor test result since last visit (if done)
• All product brands used since last visit
  (patients may have used more than one product
  during the year)

10
Cohort Selection

• To test the feasibility of monitoring inhibitor
  incidence in UDC, we studied a group of frequent
  HTC users
• Of 9,453 males with hemophilia 1,490 (16) had 4
  UDC visits
• Demographic and clinical characteristics of this
  cohort were similar to those of all UDC
  participants
• We excluded patients with lt 2 inhibitor measures
  or who had an inhibitor in the past

11
Study Methods

• Patients whose measures indicated the new onset
  of inhibitor were investigated
• Only cases confirmed by HTC staff were used
• Incidence calculated as cases/follow-up time
• Analysis used life table methods
• Relative risks were used to assess the
  statistical significance of rate differences
  between levels of risk factors and product brands

12
Findings

• 919 (75) of the 1,224 cohort members with
  hemophilia A were eligible for study
• There were 81 (8.8) prevalent cases
• The remaining 838 persons had an average of 3.9
  years of follow-up (total 3,271 PY)
• 9 new verified inhibitor cases occurred for an
  overall rate of 2.75 cases per 1,000 PY

13
Statistical Comparison of RatesAcross Factor
Product Brands
Cases per 1,000 PY
14
PTP studiesSummary
Incidence of inhibitors(ITT)
De novo inhibitors
N
()
95 CI
Product
Kogenate1 Kogenate FS2 Recombinate3 ReFacto1,4,5 A
dvate1
86 71 69 113 103
0 0 0 0 1
2 1 2 1 1
(2.3) (1.4) (2.9) (0.9) (1.0)
0.28, 8.15 0.04, 7.6 0.35, 10.1 0.02, 4.83 0.02,
5.4
CI, confidence intervals ITT, intention to treat
population
Hay 2005
1Lee Roth 2005 2Abshire et al 2000 3White et
al 1997 4Lusher Roth 2005 5Lusher et al 2003
15
Summary

• Incidence of inhibitors in PTPs is very low
• Power to detect differences in rates between
  levels of risk factors is low
• Despite nearly 4 years of follow-up on close to
  1,000 individuals, power was inadequate to assess
  differences in rates between products
• Most feasible way to study this rare adverse
  event is through post-marketing surveillance of a
  large population

16
FDA Requests

• FDA Workshop on Clinical Study Design
• Inhibitor titers done at a central lab
• Number of exposure days should be captured
• Genomic screening of enrollees
• Any product switch should be recorded
• Any lack of effect (defined as increase in dose
  or frequency of dosing) should be recorded
• International collaboration

17
Pilot Study of Post-Marketing Surveillance for
Inhibitors

• Data coordinator support for factor exposure data
  and to ensure annual blood testing
• Explore methodologies to collect required
  treatment and bleeds information
• Centralized inhibitor testing at CDC
• Hemophilia gene sequencing at CDC
• 10 HTCs selected for pilot each to enroll 50
  patients in the next year

18
Conclusions

• Large population monitoring provides the most
  power for examining rare events
• UDC is widely accepted by patients and HTC staff
  in the United States
• However, international collaboration will be
  required for adequate data to identify risk
  factors and assess product risk