Abstract
Multiple approaches are being employed to evade the cellular immune responses to adeno-associated virus (AAV) gene therapy, predominantly the use of licensed immunosuppressive medications. Corticosteroids are the most commonly used immunosuppression (IS) agents in liver-targeted AAV gene therapy trials. Some clinical studies have shown that rapid administration of glucocorticoids, such as dexamethasone and prednisone, prevented loss of transgene expression. 1 –3 The drug regimen, dose, timing, and duration of administration are all important parameters that contribute to the treatment efficacy and need to be explored further for designing immunomodulation regimens in future.
Although the administration of glucocorticoids at later time points post-AAV vector delivery has been assessed in hemophilia B clinical trials, it has not been studied extensively in animal models. A recent study in Human Gene Therapy by Chai et al. reports that dexamethasone transiently enhanced transgene expression in both wild-type and hemophilia B mice models several weeks post-AAV delivery. This increase in transgene expression was reported upon dexamethasone readministration (three or seven times).
The authors suggest that the transient increase might be clinically relevant in hemophilia patients experiencing spontaneous bleeding outcomes. When such an event occurs, transient use of dexamethasone may restore the clotting factor IX expression in patients. However, the transient elevation of transgene expression was not observed in hemophilia dog models treated with dexamethasone 1-year post-AAV delivery. This variability highlights that the rational use of steroids as well as the timing of IS administration in clinically informative in vivo models needs to be investigated.
Although several AAV clinical trials for both hemophilia A and B have had notable clinical success, an incidence of capsid-specific immune response occurred in ∼60% of patients between 4 and 12 weeks post-AAV delivery. 2,4 This was characterized by a decreased expression of Factor IX (FIX), a moderate increase in hepatocyte-derived serum alanine aminotransaminase (ALT) levels, and, in some cases, the presence of AAV capsid-specific cytotoxic T cells. 5 In early clinical trials before the usage of glucocorticoids, loss of FIX expression and an increase in transaminases were reported. 6 In subsequent hemophilia B trials, rescue steroid protocols were used if ALT rose.
Similarly, transaminitis observed in other hemophilia B AAV clinical trials was resolved by steroid treatment, as were capsid-specific T cell responses. 3,7 In a spinal muscle atrophy clinical trial and postmarketing analysis, although ∼55% of patients had elevated liver enzymes at baseline, 90% of patients showed elevated enzymes during postdosing liver functional tests. 8 Prophylactic prednisolone was administered to all patients, with differences in the dosage and duration based on the clinical status. Also, according to the recommendation, tapering would not start until aminotransferase aspartate transaminase/ALT concentrations were less than two times the upper limit of normal. 9 Liver failure was resolved in two patients with a transient high-dose steroid. 8,10
The underlying mechanism of dexamethasone-mediated increase in FIX expression after systemic AAV delivery has not been delineated yet. In general, glucocorticoids abrogate inflammatory responses by suppressing the expression of proinflammatory cytokines, such as tumor necrosis factor-alpha (TNF-α), interleukin-6, and interleukin-1-beta. 11 Consistent with earlier findings, Chai et al. found that dexamethasone treatment inhibited interferon-beta and TNF-α expression levels in the AAV-transduced mouse liver.
The authors suggest that the dexamethasone-facilitated increase in transgene expression post-AAV injection is likely due to suppression of an innate immune response and downregulation of proinflammatory cytokine production. However, how the suppression of these cytokines impact FIX protein levels needs evaluation. No changes in the AAV genome copy number or transgene expression at the transcript level were observed in the AAV-transduced hepatocytes upon dexamethasone treatment. Further understanding of how dexamethasone increases FIX protein levels could be beneficial, and additional assessments, particularly studying antitransgene antibodies, might be informative in this context.
The timing of steroid treatment might play a role in regulating the anticapsid immune response and the induction of tolerance to AAV capsid and transgene. Coadministration of a T cell-directed IS using rabbit antithymocyte globulin along with AAV-FIX vector in nonhuman primates (NHPs) resulted in the presence of anti-FIX antibodies, but none of the animals that received a later dose of IS (5 weeks post-AAV injection) developed an anti-FIX response. 12
This implies that early intensive T cell IS might abrogate the T cell and regulatory T cell (Treg)-dependent processes that develop immune tolerance, whereas a delayed IS might spare the early immune processes of Treg expansion, which thereby promotes immune tolerance induction to human FIX. Tregs induction within a few weeks post-AAV delivery observed in clinical studies for lipoprotein lipase-deficient patients and in hemophilia dogs 13 –15 supports this possibility.
The authors suggest that the enhancement of AAV transduction by glucocorticoid administration appears to vary across species. Chai et al. demonstrated that dexamethasone treatment of hemophilia B dogs subjected to systemic injection of AAV9/canine FIX vectors did not have any impact on the canine FIX transgene expression when treated 1 year after AAV delivery. Although there is evidence of differential regulation of certain genes by glucocorticoids in a species-specific manner, 16 however, the timing of initiation of IS treatment is likely responsible for the lack of response. It is important to note that there are contrasting reports regarding the IS regimen observed in a Duchenne muscular dystrophy canine model.
On one hand, sustained expression of the therapeutic canine microdystrophin gene was observed until 22 weeks in young adult dystrophic dogs injected with tyrosine mutant-AAV6 capsid after IS with cyclosporine and mycophenolate mofetil was administered for 5 weeks, starting at 1 week before AAV gene delivery. 17,18 The same transient IS regimen applied in young adult dystrophic dogs showed successful AAV gene transduction after systemic delivery of AAV9. 19 On the other hand, sustained dystrophin gene expression without any IS was observed in dystrophic dogs that received different AAV capsids through different routes. 20 –22 These studies suggest that AAV serotypes, transgene product, and route of administration are factors that also determine the parameters of IS treatment.
The variabilities already discussed may also be attributed to the lack of clinically informative in vivo models to evaluate anti-AAV immune cellular response and the associated IS treatment. To this end, the effect of prophylactic IS has been evaluated in NHPs. 12,23 Owing to a greater similarity with the human immune system, using NHPs is more advantageous for obtaining equivalent transgene expression levels, as compared with the hemophilia mouse model. In an investigational AAV gene therapy study for ornithine transcarbamylase deficiency, NHPs that received prednisolone pre- and post-AAV injection exhibited reduced interferon gamma, which corresponded to enhanced transgene expression.
Overall, more comprehensive IS approaches are required to prevent anticapsid immune responses that block transgene expression. Further studies to investigate their mechanism of immunotolerance, optimal timing, and dose of administration, and other relevant considerations will provide clinically informative information for designing immunomodulatory treatments.