Limb Perfusion Delivery of a rAAV1 Alpha-1 Antitrypsin Vector in Non-Human Primates Is Safe but Insufficient for Therapy

Abstract

Background/objectives: α-1 antitrypsin (AAT) deficiency is an inherited, genetic condition characterized by reduced serum levels of AAT and increased risk of developing emphysema and liver disease. AAT is normally synthesized primarily in the liver, but muscle-targeting with a recombinant adeno-associated virus (rAAV) vector for α-1 antitrypsin (AAT) gene therapy has been used to minimize liver exposure to the virus and hepatotoxicity. Clinical trials of direct intramuscular (IM) administration of rAAV1-hAAT have demonstrated its overall safety and transgene expression for 5 years. However, the failure to reach the therapeutic target level after 100 large-volume (1.5 mL) IM injections of maximally concentrated vector led us to pursue a muscle-targeting approach using isolated limb perfusion. This targets the rAAV to a greater muscle mass and allows for a higher total volume (and thereby a higher dose) than is tolerable by multiple direct IM injections. Limb perfusion has been shown to be feasible in non-human primates using the rAAV1 serotype and a ubiquitous promoter expressing an epitope-tagged AAT matched to the host species.

Methods: In this study, we performed a biodistribution and preclinical safety study in non-human primates with a clinical candidate rAAV1-human AAT (hAAT) vector at doses ranging from 3.0 × 10¹² to 1.3 × 10¹³ vg/kg, bracketing those used in our clinical trials.

Results: We found that limb perfusion delivery of rAAV1-hAAT was safe and showed a biodistribution pattern similar to previous studies. However, serum levels of AAT obtained with high-dose limb perfusion still reached only ~50% of the target serum levels.

Conclusions: Our results suggest that clinically effective AAT gene therapy may ultimately require delivery at doses between 3.5 × 10¹³-1 × 10¹⁴ vg/kg, which is within the dose range used for approved rAAV gene therapies. Muscle-targeting strategies could be incorporated when delivering systemic administration of high-dose rAAV gene therapies to increase transduction of muscle tissues and reduce the burden on the liver, especially in diseases that can present with hepatotoxicity such as AAT deficiency.

Keywords: AATD gene therapy; alpha-1 anti-trypsin; biodistribution; pre-clinical study; rAAV gene therapy; rAAV safety.

Figure 1

Maps of AAV vector constructs. Linear maps of the AAV gene cassettes used for this study. AAV2-ITR = inverted terminal repeat sequence from AAV2, CMVe = CMV immediate early enhancer, ACTpro = chicken β-actin promoter, Hybrid IVS = hybrid intron with upstream portion from chicken β-actin and downstream portion from rabbit β-globin, hSERPINA1= human SERPINA1 coding sequence, rhSERPINA1-myc = rhesus SERPINA1 coding sequence with c-myc epitope tag fusion, pA = SV40 polyadenylation signal.

Figure 2

Vector genome biodistribution at injection site and muscles. Vector biodistribution was measured in the rectus femoris muscle taken from a biopsy at post-injection day 28 and at the injection site as well as in ipsilateral and contralateral gastrocnemius and rectus femoris muscles at post-injection day 90 from tissue samples collected during necropsy. Results are expressed as genome copies per diploid genome. Individual results are shown and lines indicate means.

Figure 3

Vector genome biodistribution in tissues. Vector biodistribution was measured in the lymph node, lung, spleen, heart, liver, and gonads at post-injection day 90 from tissue samples collected during necropsy. Results are expressed as the number of vector genome copies per diploid genome. Individual results are shown and lines indicate means.

Figure 4

AAT serum levels. ELISA assays were performed on serum samples from NHPs that received the vector carrying the human AAT gene (rAAV1-CB-hAAT), the c-myc-AAT gene (rAAV1-CB-rhAATmyc), or vehicle control. Expression levels of hAAT or c-myc were quantified from serum samples collected prior to vector injection (day 0) and at 14, 30, 45, 60, 75, and 90 days post-injection. Data are shown as mean ± standard error of the mean. Two technical replicates were performed of the hAAT samples and four technical replicates were performed of the c-myc samples.

Figure 5

Immunohistochemical staining of muscle biopsy tissue. Skeletal muscle was collected by biopsy on day 28 post-injection, sections were immunohistochemically stained for hAAT, and the positive staining was quantified as a percentage of the total area. Tissue sections and quantification results are shown. Samples from animal 2002 were not able to be processed and the result is indicated as N/A. Three tissue sections from animal 3002 (mean ± standard error of the mean are shown) and one tissue section from the other animals were analyzed.

Figure 6

Immunohistochemical staining of muscle tissue sections. Skeletal muscle tissue was collected during necropsy at day 90 post-injections, sections were stained immunohistochemically stained for hAAT, and the positive staining was quantified as a percentage of the total area. Representative tissue sections and quantification results are shown. Two tissue sections were analyzed from each animal.