Cellular Immune Responses To Vector In a Gene Therapy Trial For Hemophilia B Using An AAV8 Self-Complementary Factor IX Vector

In a clinical study of gene transfer for hemophilia B an adeno-associated virus vector serotype 8 (AAV8) expressing a self-complementary liver-specific expression cassette for the factor IX (FIX) transgene was administered intravenously in ten affected subjects. The results of the first part of the study have been published (NEJM 365:2357-65, 2011).

In this abstract we present the immunomonitoring data, using Interferon-gamma (IFN-γ) ELISpot and polyfunctional T cell analysis of peripheral blood mononuclear cells (PBMCs) to monitor cellular immune responses to vector capsid and to Factor IX.

We have previously shown that the cellular immune response was directed solely towards AAV capsid epitopes, not FIX, and that the response was dose-dependent. Out of six subjects infused in the high dose cohort (2x10¹²vg/kg), 4/6 manifested a minor rise in liver enzyme levels and detection of capsid-specific T cell reactivitiy in the ELISpot assay at ∼7-10 weeks post vector infusion. Maximum results on IFN- γ ELISpots ranged from 200-500 sfu/million cells. In two of these cases a modest decline in FIX level also occurred. Prompt initiation of prednisolone reversed these effects and rescued FIX levels.

The remaining two subjects infused at the high dose, showed no rise in liver enzyme levels at any time point. However capsid reactive T cells were detectable in one subject as early as one to two weeks after vector infusion in peripheral blood by IFN-γ ELISpot assay, while no activation at all was detected in the other subject, possibly due to low cell recovery and viability of the cells. A similar immune response profile, with early detection of activated T cells but no rise in liver enzymes, was also observed in both subjects in the intermediate dose cohort in the first part of this study.

Polyfunctional T cell analysis revealed concurrent Interleukin-2, Tumor necrosis factor-alpha and CD107a positivity in activated T cells at the peak of activation. Furthermore it showed that capsid-specific early T cell responses were detectable in the CD4⁺ T cell and later in the CD8⁺T cell compartment.

Long-term immune monitoring of all subjects is ongoing. Importantly in one of the first two subjects treated at the high dose, capsid reactive T cells were detected by ELISpot 1.5 years after gene transfer; these cells were not detected in the other subject in whom long-term follow-up samples are available. Of note, capsid-reactive T cells were also seen at late time points (>1 year after infusion) in a middle dose subject and a low dose subject. Despite detectable T cell reactivity towards the AAV capsid in the peripheral blood FIX expression remained stable, suggesting that there is a short window of time during which transduced hepatocytes present a target for cytotoxic T cells, and that T cell positivity after this window is without any clinical consequences.

In conclusion, for this scAAV8 vector there appears to be a critical threshold vector dose for a clinically detectable immune response, starting at 2x10¹² vg/kg. The clinically detectable response occurred in four out of six subjects so far, and was manifest within a critical time interval of 7-10 weeks post infusion. The capsid-specific response was polyfunctional and detected in CD4⁺ and CD8⁺T cells in peripheral blood.

It is important to note that not all subjects treated at the high dose developed an immune response. However, given the limited dataset, it is not yet possible to define predictive parameters, e.g. HLA type of a subject, for an immune response.

Continued monitoring and future studies with more subjects will be necessary to confirm the presented findings, in particular time and rate of occurrence of a cellular response as well as successful treatment with a short course of Prednisolon.