Abstract
Introduction: Adeno-associated virus (AAV) is a versatile gene therapy vector utilized widely in clinical gene transfer studies. AAV serotype 8 (AAV8) has been used successfully as a vector in liver-directed gene transfer in hemophilia clinical trials, and a recombinant AAV8 vector is used currently in Shire's BAX 888 clinical study (NCT03370172) for the treatment of patients with severe hemophilia A. However, 1 recognized limitation of intravenous administration of recombinant AAV is the barrier posed by preexisting immunity against AAV and the cross-reactivity of anti-AAV antibodies between various serotypes that results from natural exposure to the wild-type viruses. Neutralizing antibodies (NAbs) resulting from humoral immune responses can compromise vector transduction, and reactivation of memory T cells from cell-mediated immune responses can abrogate transgene expression and eliminate transduced cells. Understanding the prevalence of cell-mediated immune responses in the hemophilia population and the association between cell-mediated and humoral immunity is important in developing effective gene therapies.
Objectives: To determine the prevalence of preexisting immunity towards AAV8 and AAV2 in the adult population of patients with hemophilia A and B, and to characterize the association between cell-mediated immunity and humoral immunity.
Methods: In this ongoing study, peripheral blood samples were collected at outpatient visits from adult males (age 18-75 years) with severe hemophilia A (plasma factor VIII [FVIII] activity <1%) or hemophilia B (plasma factor IX [FIX] activity ≤2%) for determination of immunity towards AAV8 and AAV2. Patients could elect to participate in a single baseline visit or in the longitudinal study arm for up to 3 annual visits. Patients with inhibitors or a history of inhibitors to FVIII or FIX, an underlying lymphocyte disorder, or currently receiving cytotoxic chemotherapy, monoclonal antibodies, immunoglobulins, immunosuppression, or systemic antiviral therapy were excluded. The cell-mediated immune response against AAV8 peptide antigens was quantified using an interferon-γ enzyme-linked ImmunoSpot (ELISPOT) assay to determine the number of T cells responsive to 3 different pools of AAV8 capsid-specific protein fragments. Each AAV8 antigen pool consisted of 50-60 nonoverlapping peptides composed of 15 amino acid residues. Results were considered positive if the signal was >3 times above the background and >60 cells/million. Serum was collected for assessment of humoral immunity (AAV8 NAbs or AAV2 NAbs) using a cell-based transduction inhibition assay (titer cutoff: ≥1:5).
Results: A total of 41 patients have enrolled in the study to date (mean ± SD age: 33.8 ± 11.1 years) and ELISPOT data are available for 37 patients. Of those with available data, 9/37 patients (24.3%) had a positive ELISPOT to ≥1 of the 3 AAV8 peptide pools tested. One patient tested positive for peptide pool 1, 3 for peptide pool 2, and 7 for peptide pool 3 (2 patients [5.6%] tested positive for both pools 2 and 3). Of the 9 patients with positive AAV8-specific ELISPOT results, 6 (66.7%) also had a positive result for AAV8 NAbs, while of 27 patients with negative ELISPOT results, 11 (40.7%) had a positive result for AAV8 NAbs.
Conclusions: The results from this ongoing international epidemiology study demonstrate that the prevalence of preexisting cell-mediated immunity to AAV8 resulting from natural exposure to wild-type AAV is approximately 1 in 4 adult males with severe hemophilia A or B. Furthermore, there appears to be a correlation between cell-mediated and humoral immune responses, as two thirds of patients with positive AAV8-specific ELISPOT results also tested positive for AAV8 NAbs. Understanding the relationship between cell-mediated and humoral immunity and how it may affect an individual's response to AAV-based gene therapy will be key to successfully developing this novel therapy.
Rajavel:Shire: Employment, Equity Ownership. Chapin:Shire: Employment, Equity Ownership. Tang:Shire: Employment.
Author notes
Asterisk with author names denotes non-ASH members.