Improving Viral Vector Manufacturing: Modeling Costs to Help Optimize Processes

View PDF

Manufacturing costs remain high for gene therapies delivered by adenoassociated virus (AAV) vectors. The biopharmaceutical industry must minimize such expenses because they account for significant proportions of the high prices that patients pay for treatment. During a June 2022 webinar, Emmanuelle Cameau (leader for cell and gene therapy strategic technology partnerships at Pall Corporation) joined Maxime Dumont (cell and gene therapy product manager at webcast sponsor Polyplus-transfection) to describe their companies’ efforts to model AAV manufacturing costs. Cameau and Dumont focused on productivity improvements that can be achieved by optimizing transfection reagents.

The Presentations
Dumont explained that capital expenditure (CapEx) and operational expenses (OpEx) represent 60–70% and 30–40% of AAV manufacturing costs, respectively. Among OpEx, upstream labor and materials account for a larger proportion of cost of goods sold (CoGS) than do downstream materials, with plasmids incurring the highest consumables costs (30–40% of upstream CoGS). Changing the transfection reagent can increase the productivity and reduce the amount of plasmid DNA needed per batch. Such improvements also could enable manufacturers to increase the number of doses produced per batch, decrease the requisite number of batches, or reduce batch sizes — all of which can help to control costs.

Dumont presented data from an internal study comparing the transfection efficiency of 2,000-L suspension-based processes using three different reagents: standard polyethylenimine (PEI) and Polyplus’s PEIpro and FectoVIR-AAV products. The study assumed that one dose of a gene therapy contains 8.8 × 1014 viral genomes (vg). Applying the standard PEI reagent yielded 7.8 × 1010 vg/mL. Thus, one batch generated material for 54 doses. Processes using the PEIpro and FectoVIR-AAV reagents yielded 2.45 × 1011 vg/mL and 4.8 × 1011 vg/mL, producing 168 and 328 doses per batch, respectively. Ultimately, the FectoVIR-AAV reagent increased productivity by two- and six-fold compared with what the PEIpro and PEI reagents could achieve.

Cameau bolstered Dumont’s remarks with results from experiments conducted by Pall, Polyplus, and InVitria. Scientists evaluated different reagents during transfection of a green fluorescent protein (GFP)–encoding plasmid into human embryonic kidney (HEK)293T cells. Transfections were performed in Pall iCELLis Nano bioreactors. Some processes used liposomal reagents; other processes used PEIpro or FectoVIR-AAV reagents supplemented with InVitria Optiferrin recombinant human transferrin. Sampled cells were lysed, and their contents were analyzed for fluorescent intensity to determine transfection efficiency. Processes using the FectoVIR-AAV and Optiferrin reagents showed double the productivity of those based on PEIpro and Optiferrin products — and still greater productivity than those using the liposomal reagent. Such results, Cameau explained, underscore the importance of reagent screening.

As in Polyplus’s analysis, plasmid use consistently ranked among the most critical cost drivers in Pall’s models. Cameau said that transfection processes must be optimized to minimize plasmid use. However, manufacturers also should consider downstream implications. Optimized transfection can increase levels of process- and product-related impurities, complicating subsequent purification and filtration steps. Upstream adjustments must decrease CoGS enough to overcome expenses for new downstream requirements.

Dumont listed other considerations for AAV manufacturing costs. For instance, developers hoping to manufacture products internally must evaluate how cost-effectively their facilities can meet their manufacturing goals. Consumables should be selected to provide synergistic benefits. Bioprocess equipment must enhance scalability and productivity while raising minimal concerns for product loss downstream. Implementing effective analytical technologies can enhance process monitoring and control. And therapy manufacturers should select bioprocess suppliers that can provide strong scientific and regulatory support as well as ample documentation for raw-material traceability.

Questions and Answers
What types of software can be used to model AAV manufacturing costs? Although programs such as Microsoft Excel can work, they are highly manual and introduce room for errors. Many industry suppliers have developed specialized tools and information technology infrastructures to analyze facility and process data. Suppliers also can provide valuable expertise when validating cost models.

How sensitive is CoGS to changes in price for cell-culture media? Polyplus’s and Pall’s models show that differences in media prices do not make significant impacts on CoGS. The metric is much more sensitive to plasmid price and use.

Find the full webinar online at