By Gareth Macdonald
Recombinant adeno-associated virus (rAAV) vectors are a cornerstone of genetic medicine. These hollow virus particles are used to deliver commercial gene therapies like Glybera and Luxturna, as well as many other candidates still in development.
Despite this, the biopharmaceutical industry has yet to find the most effective way of making rAAV vectors at scale. McKinsey analysts recently described vector production as one of the major challenges faced by gene therapy developers. And the situation is only going to get worse as the gene therapy market evolves and begins developing products for larger patient populations.
“With a shift beyond ultra-rare indications, viral-vector manufacturing requires rapid expansion to be able to address various diseases in the commercial space,” the analysts wrote, adding, “The broader application of viral vector-based gene therapies (for example, to more common diseases) requires higher yields and lower cost of goods (COGs).”
Soft sensing is a potential solution according a new study by researchers at the University of Ottawa and the Digital Technologies Research Center at Canada’s National Research Council, which suggests this type of monitoring and modeling can increase yields and make production cheaper.
“Soft sensors use physical or machine learning models that predict or estimate quantities of interest based on raw data. They consist of mathematical models that use data from a measurement device to predict a quantity of interest, which is used to generate new information about the bioprocess.”
The authors say while biopharma is investing millions in serum-free suspension, stable cell lines, helper-dependent AAV systems, and other methods like baculovirus-insect cells to improve yields and vector quality, use of cheaper soft sensing has been limited. And this, they argue, is a missed opportunity given that soft sensing is cheaper and more impactful than other vector yield boosting efforts.
“A typical production run of an rAAV-vector treatment utilizing high-yield cell lines and large-capacity bioreactors may only produce ten doses of a systemic gene therapy from a single batch at the cost of about $100,000 per dosage.
“Consequently, any technological advancement that decreases the cost per dose would be immediately advantageous. Soft sensors provide a fast and low-cost approach to control and monitor the bioprocesses by enabling real-time screening and control of critical process variables of the upstream process to ensure product quality and process consistency.”
Increased use of soft sensors also fits with growing biopharmaceutical industry interest in digital technologies and real-time processes monitoring, according to the authors.
“Soft sensors are a crucial enabler of Biopharma 4.0 since it enables real-time monitoring of variables that are difficult to measure directly or that can only be measured at a low sampling frequency. This is a vital step towards the digital transformation of bioprocesses.
“The application of soft sensors in upstream rAAV production can contribute to the reduction of time and production costs, the generation of relevant data, improved control of bioprocesses, increased product quality, and the reduction of raw material waste.”