Preclinical Evaluation of an Engineered AAV Capsid in Non-Human Primates for the Treatment of Haemophilia B

Introduction

Over recent years, proof-of-concept academic and early phase commercial trials have demonstrated long-term therapeutic benefit of liver directed AAV-mediated gene transfer of human FIX (hFIX). Extending this success to a greater number of haemophilia B patients is hampered by the need for large vector doses required for efficient gene transfer and associated liver toxicity. To improve potency, we focused on developing a novel synthetic capsid, AAVS3, to increase human hepatocyte transduction.

Here we present the preclinical in vitro and in vivo evaluation of AAVS3 pseudotyped vectors. In addition to showing strong liver tropism we demonstrate clinically relevant expression levels of hFIX expression in non human primates.

Methods

Primary human hepatocyte cultures were used to compare transduction efficiency of our novel capsid with AAV5 and AAV8, variants currently in the clinic. To extend this data in vivo, a humanised mouse liver model was utilised. Subsequently, fourteen Rhesus macaques were dosed with AAVS3 pseudotyped FIX vectors, via saphenous vein injection. Levels of hFIX were determined using an ELISA assay that is able to distinguish hFIX protein from the native Rhesus macaque FIX protein. At harvest biodistribution of vector was detected by quantitative PCR.

Results

We demonstrate that gene transfer efficiency with AAVS3 was up to 10-fold higher in primary human hepatocytes, when compared to vectors pseudotyped with AAV serotype 5 or 8 capsids. In addition, studies in a chimeric human-murine liver model revealed AAVS3 was 6-fold more efficient at transducing human hepatocytes compared with an AAV8 vector. Using two distinct optimised FIX expression cassettes, we demonstrate highly efficient gene transfer in Rhesus macaques. Indeed, AAVS3 FIX vectors injected at clinically relevant doses of 0.93 x10¹² to 2.56 x10¹³ vg/kg achieved mean peak levels of 0.546 and 21.7µg/ml of hFIX respectively (~11% to ~430% of normal levels). Biodistribution studies showed strong liver tropism following peripheral vein administration with minimal distribution to other organs at or below the level of detection. No vector associated lesions or adverse events were observed in the liver.

Conclusion

Collectively, these data support further evaluation of an AAVS3 pseudotyped vector in humans with the aim of achieving transduction of a larger proportion of hepatocytes using a lower, potentially safer dose of vector than achieved with AAV8.