The utilization of recombinant adeno-associated virus (rAAV) in gene therapy delivery remains highly effective to treat a range of diseases and conditions, evidenced by numerous ongoing clinical trials and recent regulatory approvals.
The industry has adopted standardized methods for the production and purification of rAAV, streamlining processes for greater efficiency. However, high titers are crucial for meeting demand and cutting costs for therapies that require large doses or have wide patient bases.
This poster presents process considerations in the optimization of two manufacturing case scenarios with different titers.
Download this poster to explore:
- An affinity capture chromatography method in clinical and commercial manufacturing contexts
- How to leverage dynamic binding capacity (DBC) data to estimate optimal productivity of rAAVs
- Process conditions and column geometries to meet maximum processing times while maximizing resin utilization
Productivity optimization and process calculations for AAV affinity chromatography INTRODUCTION The use of recombinant adeno-associated virus (rAAV) as a delivery method for gene therapies continues to be successful with hundreds of ongoing clinical trials and some recent approvals. The diversity of applications for rAAV ranges from rare diseases affecting small patient populations to more prevalent inherited ailments such as hemophilia. The doses required vary widely from ~4E11 vg/eye for subretinal administration to 3.5E14 vg for intrathecal applications [1]. From a manufacturing perspective the field has moved to common approaches for production and purification of rAAV. Upstream approaches typically use transfection of HEK293 cells and titers are routinely in the 1-2E10 vg/mL although higher titers of up to 6E11 vg/mL at a 2000 L scale were recently reported [2]. These high titers will be needed for large dose and/or patient populations to meet the demand of these therapies and reduce costs. For rAAV purification the majority of the field has moved to scalable processes employing an affinity capture chromatography step [3] and commonly utilizing POROS™ CaptureSelect™ AAVX resin. In this work, dynamic binding capacity (DBC) data for multiple AAV serotypes were leveraged to estimate an optimal productivity of rAAV using the AAVX resin. An analysis of process conditions and column geometries that would fit maximum processing times and resin utilization was conducted for two case scenarios representing current titers for clinical manufacturing and high titers for commercial manufacturing scales. Alejandro Becerra-Arteaga, Ph.D. and Jett Appel, Thermo Fisher Scientific, Bedford, MA, USA Dynamic Binding Capacity ✓ Limited DBC data are available due to high capacity of AAVX resin, relatively low titers, and sample availability. ✓DBC for AAV2 is relatively high (~1E15 capsids/mL resin) even at 30 sec residence time. ✓Data fit to equation (I) approximates the dependence of DBC to residence time. TRADEMARKS/LICENSING © 2023 Thermo Fisher Scientific Inc. All rights reserved. All trademarks are the property of Thermo Fisher Scientific and its subsidiaries unless otherwise specified. This information is not intended to encourage use of these products in any manner that might infringe the intellectual property rights of others. Intended use: For research use only. Not for use in diagnostic procedures CONCLUSIONS • The relatively high binding capacity of POROS CaptureSelect AAVX resin was confirmed to be >1E15 capsids/mL resin at residence times >= 0.5 min for AAV2. • Productivity is maximized at load residence times <= 0.5 min depending on titer but hardware and/or system considerations limit operation closer to 1 min. • For clinical manufacturing the high DBC allows for a range of process conditions and requires small column volumes. • For large bioreactor volumes and high titers the model suggests columns 20-30 cm diameter to meet typical processing limits while maximizing resin utilization. Scale-up and process considerations Productivity ✓Productivity maximum is achieved at residence times below 0.5 min. ✓Productivity increases by ~3.5x with an increase in titer of ~12x. ✓ Increased titer shifts productivity maximum from ~7 to ~24 seconds RT for loading. METHODOLOGY Dynamic binding capacity: AAV2 breakthrough curves were generated using HEK293 clarified lysate to determine DBC at 10% breakthrough. AAV8 and rh10 DBC data were obtained from references 4 and 5, respectively. Equation I was fitted to the DBC data using a linear regression numerical method. Productivity: Productivity curves were generated using equations I and II. Column volumes and residence time for the non-loading steps were 25 CV and 2 min. Column volumes and residence time for CIP steps were 10 CV and 3 min. Scale-up and process considerations: GMP pre-packed column pressure limitations were based on literature from multiple vendors. Pressure drop at 3 bar was based on pressure-flow curves for POROS CaptureSelect AAVX resin (internal pressure-flow data). Case scenarios Processing time and resin utilization calculations were performed using Microsoft ® Excel ® assuming 20% full capsids and the results were further analyzed and plotted using MODDE® software. Scenario 1. Clinical manufacturing, 200 L, Co=2.5E11 capsids/mL 1.0E+12 1.0E+13 1.0E+14 1.0E+15 1.0E+16 0 1 2 3 Dynamic Binding Capacity @10% breakthrough (capsids/mL resin) Residence time (min) AAV2 AAV8 rh10 Fit Qd= Dynamic Binding Capacity @10% breakthrough Qdmax = DBC at long residence times RT = Load residence time q = Residence time constant 𝑃= 𝐴𝑚𝑜𝑢𝑛𝑡 𝑜𝑓𝐴𝐴𝑉 𝑐𝑎𝑝𝑠𝑖𝑑𝑠 𝑝𝑢𝑟𝑖𝑓𝑖𝑒𝑑 𝑈𝑛𝑖𝑡 𝑟𝑒𝑠𝑖𝑛 𝑣𝑜𝑙𝑢𝑚𝑒 ×𝑈𝑛𝑖𝑡 𝑡𝑖𝑚𝑒 P = Productivity h = Loading safety factor (% DBC) C0 = Load sample concentration CVnon-load= Column volumes for non-loading steps RTnon-load = Residence time for non-loading steps CVCIP= Column volumes for non-loading steps RTCIP = Residence time for non-loading steps No breakthrough 0 500 1000
