614. Frequency of T Cell Responses to AAV-2 Capsid in Peripheral Blood Mononuclear Cells of Normal Subjects Does Not Correlate with Anti-AAV2 Antibody Titers

Abstract

Top of pageAbstract Pre-existing immunity to AAV2 may affect AAV-mediated gene transfer by blocking transduction or by causing destruction of AAV-transduced cells. In a clinical trial for AAV2-mediated gene transfer for hemophilia B, subjects infused with the highest dose of vector showed that neutralizing antibody (NAB) titers of 1:2 were compatible with successful AAV2 transduction, while a titer of 1:17 blocked transduction by vector infused through the hepatic artery. In addition, detailed T cell studies in one subject showed a transiently detectable AAV2 capsid specific T cell response at weeks 2 to 8 after AAV2-hF.IX infusion into the hepatic artery (MolTher 9:S383). We hypothesized that characterizing the immune response to wild-type AAV2 (as a consequence of natural infection) in the normal population may allow us to understand the impact of pre-existing immunity to AAV on AAV2 vector mediated gene transfer. Peripheral blood mononuclear cells (PBMC) and serum were isolated from 21 normal subjects. Total anti-AAV2 IgG antibody, neutralizing AAV2 antibody titer and HLA genotype were determined. PBMC responses were measured by ELISpot assay. The total anti-AAV2 antibody titer ranged from undetectable to 8,339 ng/ml. NAB titers ranged between <1:3 to 1:316-1:1000. There is a correlation between NAB titer and total anti-AAV2 IgG. Of the 15 subjects in whom NAB were determined, 7 showed NAB |[ge]|1:17, i.e. would not have been candidates for transduction by hepatic artery infusion of vector. (Lower titers may also have prevented transduction; informative data in humans do not exist for titers between 1:2 and 1:17). To define the frequency of AAV-capsid-specific T cells, we carried out ELISpot assays to analyze IFN-|[gamma]| secretion after stimulation with AAV2 capsid derived peptides. Two of 21 normal subjects analyzed had detectable IFN-|[gamma]| secretion by PBMCs in response to AAV2 capsid derived peptides. Responses mapped to 3 distinct peptides (peptides 81 and 237 in one subject, peptides 81, 90, and 237 in the other), with at least one of these peptides predicted to be the highest scoring MHCI binding peptide for B8, the MHC genotype of one of the subjects, using the RANKPEP search algorithm. There is no correlation between antibody titers and IFN-|[gamma]| secretion by PBMCs in response to stimulation with AAV2 capsid peptides. For example, the donors with the highest NAB titers had no detectable IFN-|[gamma]| secretion by ELISpot. Therefore, the B cell response is not predictive of a detectable T cell response in the peripheral blood. We conclude: 1) the frequency of detectable T cell responses to AAV capsid in PBMCs from normal adults is low; 2) undetectable responses in PBMCs at baseline do not predict absence of AAV capsid-specific T cell responses after AAV vector infusion into the hepatic artery; 3) analysis of T cells from other compartments, e.g. the spleen (currently underway in our laboratory), may be required to define the frequency and activation status of AAV capsid-specific T cells in the normal human population. G. Pierce and M.Tigges are employees of Avigen, Inc.