When Immune Suppression Goes Wrong

Main Text

Many clinical in vivo gene therapies for genetic disease, ranging from early-stage trials to approved medicines, utilize adeno-associated viral (AAV) vectors. Nonetheless, immunogenicity remains a major obstacle. Humoral and cellular immune responses against the vector’s capsid or the transgene product can lead to a loss of therapy, prevent re-administration of the vector, and cause immune toxicities that are potentially harmful to the patient.1, 2, 3 Therefore, immunosuppression (IS) regimens are increasingly being included in the design of clinical protocols. Ironically, studies in non-human primates (NHPs) show that IS protocols can achieve the opposite and induce immune responses in AAV gene therapy, depending on the composition of the drug cocktail and timing of administration. These findings re-emphasize the importance of careful evaluation of IS regimens and pre-clinical trials in NHP models before applying the therapy to human patients.

We would like to remind readers of an earlier study in which the safety of AAV-mediated gene delivery was evaluated when combined with different sets of anti-T cell IS regimens in NHPs.⁴ The ultimate goal of the approach was to develop a protocol that could counter CD8⁺ T cell responses against AAV capsid in humans. An AAV2 vector expressing human clotting factor IX (hFIX) under the control of a liver-specific promoter was intravenously injected into 3 groups of male rhesus macaques (n = 3 in each group). The animals received no IS regimen (control group), a regimen consisting of MMF (mycophenolate mofetil), sirolimus (rapamycin), and daclizumab (3 IS group), or a regimen consisting of MMF and sirolimus (2 IS group). MMF and the mTOR inhibitor sirolimus had already been used in different settings of organ transplantation. Daclizumab is an anti-interleukin-2 (IL-2) receptor antibody, specifically binding to the CD25 subunit of the human IL-2 receptor, thereby depleting activated T cells. While these immune-suppressive agents were known to work well together in solid organ transplantation, it was unknown whether they would function in the same way in liver-directed gene transfer coupled with AAV vectors. Unexpectedly, the animals that received the IS regimen of all 3 drugs developed high levels of neutralizing antibodies against the hFIX transgene product. In contrast, the animals receiving no drugs in the control group, or only MMF and sirolimus, did not form antibodies against the transgene product. Moreover, the animals that received all 3 drugs formed a higher titer of immunoglobulin G (IgG) against AAV capsid than those animals receiving either no drug or only 2 drugs. The vector copy number analysis using needle biopsies from livers showed no difference across the animals in all groups, indicating that the efficiency of liver-directed gene transfer was not altered by the IS regimens. Interestingly, a reduction of CD4⁺CD25⁺FoxP3⁺ regulatory T cells (Tregs) was observed in the animals receiving the 3 drugs. These data suggested that addition of daclizumab to MMF and sirolimus prevented tolerance induction to the transgene product by depletion of Tregs, which constitutively express CD25. Hence, daclizumab’s effect on Tregs had more deleterious consequences than the desired effect on effector T cells.

Another important parameter to be considered for incorporation of IS into AAV-mediated gene therapy would be the time when an immune-suppressive agent is administered, as now shown in a new study by Samelson et al.⁵ in Molecular Therapy - Methods & Clinical Development. Samelson et al.⁵ assessed a potential risk of a T cell-targeted IS regimen, antithymocyte globulin (ATG), for AAV-mediated gene delivery in NHPs. ATG is an antibody derived from rabbits or horses suppressing human T lymphocytes and is being used in organ transplantation and therapy for aplastic anemia, and its efficacy in gene therapy has been unclear. Rabbit ATG was intravenously administered to male rhesus macaques (n = 3 in each group) at around the time of gene delivery with AAV2 vector expressing human FIX under a liver-specific promoter (early ATG group) or 5 weeks after vector administration (delayed ATG group). Both groups also received other T cell-directed IS agents: MMF beginning 1 week before vector administration and sirolimus starting at the time of vector administration. While 2 out of 3 animals in the early ATG group developed antibodies against the hFIX transgene product, none of the animals in the delayed ATG group formed any detectable level of anti-hFIX. Intriguingly, the animal in the early ATG group that did not develop anti-hFIX showed the highest frequency of Tregs in the group. The two animals with anti-hFIX formation showed comparatively higher ratios of Th17 cells to Tregs. In addition, all animals in the delayed ATG group that received ATG 5 weeks after vector delivery did not show T cell responses against AAV.

Altogether, these observations suggest that concurrent administration of ATG at the time of gene delivery to the liver by AAV vector may interfere with Treg-mediated tolerance induction to the transgene product. This study also illustrates a critical time window for induction of immune tolerance against transgene products, which is known to depend on Tregs. Although the main focus of the study is on immune response to the transgene product, a potential effect of ATG on humoral responses to the AAV capsid needs to be carefully considered as well. While all 3 animals in the early ATG group showed a delayed anti-AAV IgG formation after the withdrawal of the IS regimens, delayed ATG suppressed IgG production against capsid in 2 of 3 animals. For data interpretation, one has to caution that a caveat of this study (and many others in the expensive NHP model) is the low number of animals. Whether this regimen would be effective in blocking CD8⁺ T cell responses against capsid is also not known. To suppress tissue inflammation and counter CD8⁺ T cell responses, steroid drugs have found their way into several clinical AAV gene transfer protocols. As the field attempts to develop more diverse approaches to IS to counter different kinds of immune responses, careful pre-clinical evaluation remains paramount. While transient immune suppression is a promising approach to enforce lasting gene therapy, studies in NHPs illustrate that the regimen has to be carefully designed and evaluated with regard to the choice of drugs and dosing schedule. Protocols that have been successful, for example, in organ transplantation or autoimmune disease, may not directly translate to gene therapy. Importantly, wrongly designed IS may actually induce or heighten immune responses to transgene product and vector, in particular, when interfering with Tregs.