Personalized Prediction of Tolerance and Immunogenicity in Hemophilia

We are developing a novel, personalized strategy for assessing immunogenicity of protein
therapeutics using, as our model, the infusion of Factor VIII (FVIII) into hemophilia A (HA)
patients. About 20% of all treated HA patients develop neutralizing FVIII alloantibodies
("inhibitors") that make disease management difficult and expensive. Nowadays, HA is usually
treated with highly purified human recombinant (r)-proteins, an advance in safety from
pathogens not accompanied by a decrease in inhibitor incidence. Current strategies for
upcoming FVIII formulations focus largely on engineering the most immunogenic epitopes in
the hope of forming a universally less immunogenic protein. In contrast, we are pioneering a
pharmacogenetic approach to immunogenicity that takes into account the underlying
variability of the patient population.

This project focuses on defining the role of individual genetic differences on FVIII
immunogenicity. The principles, however, have broader application for protein therapeutics
in general. We have studied non-HA-causing variants in the FVIII gene (F8) and have shown
that (i) nonsynonymous (ns)-single-nucleotide polymorphisms (SNPs) encode several
structurally distinct wild-type FVIII proteins in the human population and (ii) a sequence
mismatch between patients' endogenous FVIII and infused FVIII due to ns-SNPs is a risk
factor for inhibitor development that may explain the high inhibitor incidence in HA
patients with black African ancestry.

The most well established risk factor for inhibitor development is the type of HA-causing F8
gene mutation. As a rule, large alterations in F8 and absence of antigenically
cross-reactive material (CRM) in plasma are associated with inhibitor development. The most
common F8 mutation causing severe HA, an intron-22-inversion (I22I), fits that description
but is not associated with a high inhibitor risk. Similarly, while most HA patients with
missense mutations do not develop inhibitors, this alloimmune complication occurs frequently
in patients with one of a few highly recurrent missense mutations.

While not definitively established, population heterogeneity in the repertoires of
HLA-class-II (HLA-II) molecules expressed on the surfaces of the antigen-presenting cells in
individual patients is likely another genetic contributor to inhibitor risk.

This project is a comprehensive assessment of the pharmacogenetics of the immune response to
FVIII leveraging a unique resource comprised of a group of 55 subjects with severe or
moderately-severe HA who were (i) enrolled as previously-untreated patients (PUPs) in the
recently concluded clinical trial known as the Advate PUP study and (ii) have received the
same r-FVIII protein (i.e., Advate) since birth. Prior PUP-study data as well as new blood
samples and data will be obtained from these subjects upon their enrollment into the current
study. In addition to having been treated with only a single FVIII product, this exceptional
patient cohort was (and continues to be) closely monitored for both FVIII infusion history
and inhibitor development, the latter of which by undergoing frequent Bethesda testing. (HA
patients who have been treated with several FVIII products are not ideal for testing the
hypotheses we have proposed.)

Inclusion Criteria:

- Patients with severe or moderately severe hemophilia A (HA) who have since birth been
treated with only a single Factor VIII product (i.e., FVIII protein molecules
containing only one primary amino acid sequence).

Exclusion Criteria:

- HA patients with severities other than severe or moderately severe.

- Hemophilia B patients.

- HA patients who have been treated with more than one FVIII product.

- HA patients who have been treated with more than one FVIII product.

- HA patients who do not have verifiable infusion histories.

- HA patients who lack documentable inhibitor testing & infusion histories.