Adeno-Associated Mediated Gene Transfer for Hemophilia B:8 Year Follow up and Impact of Removing "Empty Viral Particles" on Safety and Efficacy of Gene Transfer

Background: We have previously shown that a single intravenous administration of a self-complementary adeno-associated virus (scAAV) vector containing a codon-optimised factor IX gene, under control of a synthetic liver specific promoter and pseudotyped with serotype 8 capsid, (scAAV2/8-LP1-hFIXco) resulted in a dose-dependent increase in plasma FIX levels in all 10 enrolled severe hemophilia B (HB) patients (ClinicalTrials.gov:NCT00979238; Nathwani et al 2011). FIX activity was stably maintained for at least 3 years (Nathwani et al 2014) but concerns over FIX expression declining over time remain. This is because AAV-mediated transgene expression is mediated mainly by episomally retained viral genomes, which may be lost with natural hepatocyte turn-over. The only vector-associated adverse event was an asymptomatic rise in liver enzymes associated with a decline in FIX levels, occurring within 3 months of gene transfer in two-thirds of the patients treated at a dose of 2x10¹² vector genomes(vg)/kg. Liver enzymes normalized with corticosteroids without complete loss of transgene expression. There was no long-lasting toxicity over a period of 3 years but further follow-up is required. The vector preparation used contained an excess of empty capsids, which lacked a full-length viral genome, and are therefore, non-functional but capable of provoking an immune response against transduced hepatocytes. Therefore, a new clinical preparation of scAAV2/8-LP1-hFIXco was manufactured from which most of the empty particles were removed by caesium chloride density centrifugation in the hope that this would reduce the risk of hepatotoxicity. We report on the evaluation of this new vector preparation in severe HB patients and provide an update on up to 8 years follow-up of our original cohort of patients.

Methods: Ten subjects were recruited in 2010-2012 to the initial dose-escalation/extension study arm, which entailed a single intravenous infusion of scAAV2/8-LP1-hFIXco (full: empty capsid ratio ~1:10) at a dose of either 2x10¹¹vg/kg, 6x10¹¹vg/kg or 2x10¹²vg/kg. Two severe HB patients (FIX <1%) were enrolled into the first and mid-dose cohorts, with six patients treated at the high dose. In a follow-on study arm, two severe HB subjects received a dose of 2x10¹²vg/kg of the new scAAV2/8-LP1-hFIXco preparation (full: empty capsid ratio 1:3) whilst the next 2 patients were treated at a dose of 5x10¹²vg/kg. In both arms, vector was administered without prophylactic immunosuppression but corticosteroids (starting at 60mg/day) were commenced if liver enzymes increased to ≥2 fold over baseline levels after gene transfer.

Results: Transgenic FIX activity levels have remained stable in all 10 subjects treated in the initial dose escalation/extension arm over a median follow-up of 6.7±1.0 years with mean levels in the three dose cohorts at the time of reporting of 1.9±0.6, 2.3±0.3 and 5.1±1.4 IU/l respectively. Over this period, annual FIX concentrate usage has dropped by 66% and annual bleed rate has declined by 82% when compared to pre-gene therapy levels. No late toxicity was observed. Neutralising antibodies to FIX were not detected in any patient but anti-AAV8 capsid-specific antibody levels persisted at high titres in 9 of 10 patients. In patients treated with the new preparation of scAAV2/8-LP1-hFIXco (median follow up = 2.1±1.4 years), mean FIX activity in the 2x10¹²vg/kg dose cohort was 2.6±0.7 IU/l. This is lower than observed previously at this dose level, but the difference is not statistically significant. Mean steady state FIX levels in the 5x10¹²vg/kg cohort were 17±5 IU/l. FIX antigen to activity ratio was 1:1. Elevation of serum alanine aminotransferase was observed in 3 of 4 patients treated with the new vector preparation, requiring treatment with corticosteroids.

Conclusion: This is the first report to demonstrate stable therapeutic expression of FIX in patients with severe HB over a period of 8 years following systemic administration of scAAV2/8-LP1-hFIXco without late toxicities. We show for the first time that reducing the capsid load by removing empty AAV capsids does not appear to reduce the incidence of hepatotoxicity in patients with severe HB suggesting that other factors are involved in the aetiology of this process.