Sequencing-based analytics offer potential for rapid sterility and adventitious agent testing

I’m Gary Sommerville and I’ll be sharing how ElevateBio is developing sequencing-based analytics with the potential to help partners reduce release testing timelines by detecting bacterial, fungal, and viral contaminants faster and more comprehensively.

Currently, genetically engineered cell therapies take about 30 days from collection to reinfusion. For cancer patients with conditions like leukemia or multiple myeloma, reducing this vein-to-vein time is critical, and major bottleneck is traditional sterility and virus testing, which follows USP guidelines requiring growth-based detection methods. These can take up to 28 days for sterility testing and 5 to 7 days even with rapid methods. When every day matters for critically ill patients, we need faster alternatives that maintain the same level of sensitivity.

Here at Elevate, our solution is developing a next-gen sequencing approach that can simultaneously detect bacterial, fungal, and viral contaminants in cell therapy products. The workflow begins with DNA and RNA extraction from the cellular product, followed by library prep and hybrid capture enrichment. We then use Illumina short-read sequencing, coupled with a custom bioinformatics pipeline that can identify and quantify contaminating organisms within 42 to 72 hours rather than weeks.

Our proof-of-concept studies demonstrate robust detection of the five USP required test organisms. We spike DNA from these organisms, including Aspergillus brasiliense, Bacillus subtilis, Candida albicans, Pseudomonas aeruginosa, and Staphylococcus aureus into representative cell therapy products. Remarkably, we achieved detection at approximately ten colony forming units per 75,000 cells, approaching the sensitivity required by compendial standards. This represents both bacterial and fungal pathogens that are common concerns in manufacturing.

For adventitious virus testing, we use a commercially available viral reference standard, Viriome Mix. Our methods successfully detected five different viruses at extremely low levels between 1 to 10 genome copies per 7,500 cells. This demonstrates the versatility of our platform to detect both DNA and RNA viruses that could potentially contaminate cell therapy products, providing comprehensive safety assessment in a single assay.

The computational backbone of our system is a custom next flow pipeline deployable on AWS for scalability. For sterility testing, we align reads to both human and target organism genomes. For adventitious agents, we extract unmapped reads and classify them against comprehensive databases. This dual approach ensures we can both detect known targets and also discover unexpected contaminants. We’ve demonstrated initial feasibility of NGS-based methods as a potential enhancement or a replacement for traditional compendial testing.

Our approach offers several advantages simultaneous detection of multiple contaminant types, faster turnaround times, and digital quantification moving forward were focused on further characterizing and improving our limit of detection to fully meet regulatory requirements. We’re also exploring new sequencing technologies, including long-read platforms that could further decrease turnaround time and cost while maintaining sensitivity. This work represents an important step towards modernizing quality control in cell therapy manufacturing, potentially reducing vein to vein time and improving patient access to these life-saving therapies.

This work shows how NGS-based contaminant detection could help partners move towards faster, more informative release testing while maintaining the safety expectations required for complex advanced therapies.