A Broad AAV Formulation Platform for Bespoke Genomic Therapies
Purpose: Adeno associated virus (AAV) is a versatile vehicle for delivering recombinant DNA for genomic medicine (gene therapy, gene regulation, and genome editing). AAV is comprised of genetic material enclosed within a viral capsid. Using AAV, it is possible to create multiple viral vector therapeutic candidates by modifying the genetic payload enclosed within a common viral capsid. This “platform approach” can streamline many activities required for development of multiple genomic therapies by leveraging data that is capsid-specific across programs. Commonly, a single AAV formulation is used for multiple indications. However, a “one size fits all” platform formulation may not be suitable to treat the vast range of existing genetic diseases. Therapeutic dosages can vary substantially for different genomic therapies depending on various factors including the patient population (adult vs. pediatric), target tissues, mechanism of action, and route of administration. In these cases, a product formulation tailored for the target indication and population can improve patient safety and comfort. Methods: Formulations at approximately 1, 3, 5, and 10 x 1013 vector genomes/mL were evaluated in this study. Each formulation contained the same excipients – only the active ingredient concentrations were adjusted. These concentrations were selected as the most practical and feasible to meet the needs of several clinical dose ranging studies as well as maintaining high yield from manufacturing batches. All AAV in this study was produced with the Sf9 cell line in suspension bioreactors. Drug substance was purified with multi-stage column chromatography. Purified drug substance was concentrated to the highest level and buffer exchanged into the product formulation buffer by tangential flow filtration. This material was aliquoted and diluted to the target concentrations with formulation buffer. Samples were sterile filtered and manually filled into 2 mL polymeric vials, stoppered, and sealed. Several stability-predicting tests were performed on initial samples. Thermal stability was measured by monitoring sample fluorescence and light scattering over differential scanning thermal ramps. This analysis was performed to measure temperatures at which capsid denaturation (Tmelt), aggregation (Tagg), and genome leak (Tleak) occurs. Extended duration elevated temperature isothermal holds were also performed also monitoring fluorescence and light scattering. Thermodynamic and colloidal stability parameters (zeta potential, kD, and B22) were also measured on initial samples with dynamic and electrophoretic light scattering. The intended long-term storage for AAV drug products is at least 2 years at ≤-65°C with only 1 freeze/thaw cycle expected, and typical point-of-use conditions are several days at ambient temperature. To challenge these samples, up to 3 months at 25°C and up to 1 month at 40°C were selected for accelerated stability, and up to 10x freeze/thaw cycles were performed. Samples were also incubated at ≤-65°C and 2-8°C for up to 12 months. Critical quality attributes including vg titer, capsid concentration, full capsids, appearance, pH, size distribution, and aggregation were measured throughout the stability study with results compared against pre-determined acceptance criteria. Also, several characterization tests including particle concentration, subvisible particles, infectious titer, and protein purity were monitored for information only. Testing of the high and low concentration samples was prioritized to establish any trends in data. Periodic sampling of middle concentration samples was performed to verify trends were consistent. Results: Data are presented on the stability of several AAV6 product formulations over multiple concentrations and with several different genetic payloads. Stability-predicting tests showed equivalent capsid denaturation (78.6-79.1°C), aggregation (75.3-75.8°C), and genome leak (55.7-56.3°C) temperatures throughout the concentration range studied with good reproducibility (%CVs < 1.25%). Thermodynamic parameters also suggested AAV is colloidally stable in the formulation buffer and at the concentrations studied. The product quality attributes of high and low concentration samples passed the predetermined acceptance criteria when incubated at ≤-65°C, 2-8°C, 25°C, and up to 10x freeze/thaw cycles. At accelerated conditions (40°C), product quality attributes began to be affected with some loss in titer and aggregation observed. However, the trends in these quality attributes at accelerated conditions were equivalent in the high and low concentration formulations. Stability trends and observations did not differ when multiple different genetic payloads were evaluated. Conclusion: These results establish a stable platform design space for AAV products of varying active ingredient concentrations and multiple different genetic payloads. A “bespoke fit” formulation can now be selected from this design space to suit the needs of a given disease indication. The bracketed study design employed also enabled a range of stable AAV drug product formulations with reduced testing required.