ISEV2020 Summary – Part II

Whether Upstream or Downstream,  EV Manufacturing is Becoming Seaworthy in 2020

Manufacturing and Isolation Methods

After exploring the Conference’s highlights (Part I) with basic biology, applied technologies and clinical translation, and analytical methods, we now turn our attention to how ISEV2020 presented ways to cultivate large numbers of EVs, and how to concentrate and isolate them in downstream process. Both upstream and downstream processes receive appreciable attention from academic institutions and industry, which reflects the EV field’s efforts towards providing a complete solution to clinical manufacturing at a commercial scale.

RoosterBio exhibited two presentations on scalable manufacturing of hMSC-EVs. Dr. Jon Rowley (Founder & Chief Product Officer of RoosterBio) presented strategies for simplifying MSC-EV product development and shortening clinical development timelines of MSC-EVs via supply chain innovations. Dr. Taby Ahsan (Vice President of Analytical, Process & Product Development) presented the different upstream considerations for scalable manufacturing of MSC-EVs. Both emphasized on the critical need of a scalable hMSC production platform such as bioreactors, and consequently clinically relevant hMSC-EV lot size, while paying attention to the various EV productivity metrics to ensure an efficient production process that maximizes scale and minimizes COGS.

Upstream culture parameters are known to affect hMSC growth and MSC-EV production. Researchers from University of Lisbon (Raquel Cunha of Joaquim Cabral’s and Cláudia Lobato da Silva’s groups) showed that hypoxic condition facilitates better MSC growth than normoxic conditions. In addition, Professor Steven Jay at University of Maryland showed that modification of upstream culture parameters can not only increase cell yield, but also change immunomodulatory functions as shown by enhancement of anti-inflammatory characteristics through priming and dynamic culture.

It is worth noting that while MSC is a popular choice of cell type for generating therapeutic EVs, another approach to increasing upstream EV productivity is by using “younger” cells as a source, as presented by Dana Larocca from AgeX Therapeutics. Their approach of generating a variety of cells from embryonic progenitor cells can yields billions of doses of 100M cells/dose from a single cell source, and consequently large number of EV doses.  It would be interesting to see if iPSCs generated from adult or neonatal cell sources like MSCs could provide an alternative.

The EV field overall has converged on the realization on the importance of using bioreactors as a production platform to achieve clinically and commercially relevant lot sizes. Numerous studies were presented on scaleup of MSC-EV production using a variety of bioreactors: spinner-flasks, vertical-wheel bioreactors, hollow fiber bioreactors. All studies consistently showed increase in particle concentration in 3D compared to 2D flasks, with productivity reaching up to 4.4×1010 particles/ml in the FiberCell bioreactor, as presented by Dr. Laura Perin of Children’s Hospital Los Angeles. With proper manufacturing and downstream processing, such system can support clinical applications.

Downstream processing of EVs is just as important; no one wants to manufacture products at a scale only to waste a large fraction of it due to incapability to process. Like in cell therapy, efforts are being made to address downstream bottlenecks in EV production. There is a general agreement that there is no single EV downstream platform that does it all, and that selection of EV isolation method should be made based on sample type, analysis of choice, while keeping in mind the balance between recovery and specificity.

Studies on the development of downstream processes & tools showed comparison to some of the standard downstream methods such as ultracentrifugation. Researchers from Johns Hopkins University (Dr. Liang Dong from Professor Kenneth Pienta’s lab – in collaboration with Professor Kenneth Witwer) reported on the use of Exodisc, a centrifugal microfluidic nanofiltration system, to purify EVs from various fluid samples. The study highlighted the high EV yield and high purity achieved from Exodisc compared to ultracentrifugation.

Affinity-based isolation method has also gained a lot attention due to its capability to generate highly purified EVs. Fujifilm utilized affinity-based EV isolation method using Tim4 (a receptor with a strong affinity to phosphatidylserine – which is present on EV membrane), instead of the typical CD9, CD63, CD81 antibodies. Their data showed cleaner (homogeneous, non-aggregated / non-fused) EVs and higher EV markers concentration were achieved with Tim4 affinity method compared to ultracentrifugation. It was also shown that the EVs isolated by this method shows higher anti-inflammatory and anti-fibrotic activity.

The industry is moving towards streamlining the downstream processing by providing commercial kits that combines EV isolation and analytics. Izon, known for their size exclusion chromatography (SEC) columns to isolate EVs, presented various kits for EV characterization to complement their isolation product. Also under development are larger scale columns that can support processing of 5-10L of conditioned media, as well as combination strategies to further separate particles following SEC. The commercial kit format is also available for Fujifilm’s Tim4 capture antibody platform, as both ELISA and flow cytometry kits.

Further downstream, solutions for EV storage such as Fujifilm’s EV-Save™ Blocking Reagent, which suppresses adsorption of EVs to lab tools and also acts a cryoprotectant, will help EVs move through the clinical translation. Finally, CMOs like Lonza offer various capabilities from filtration, TFF, anion exchange chromatography, and combined with services in the EV characterization – to provide a complete upstream / downstream solution for EV manufacturing.

Conclusion – All Products Begin With a Process

If “heterogeneity” is a friendly word for the discovery scientist in search of novel biomarkers or diversity in EV populations, it might be an equally disquieting one for the manufacturer of EVs for clinical products, or supplier of supporting ancillary materials.  Specifically, biologics drug components like EVs need to be manufactured with good comparability across parameters of identity, safety, purity, and potency.  Due to EVs’ inherent process complexity, it may be realistic to monitor only the parameters of key relevance for safety and efficacy, [1] [2] a matter of continuous challenge and revision, with ISEV’s essential input.

Fortunately, as we learned from ISEV2020, the tide for EV therapeutics rises higher each year, lifting all kinds of new and specialized vessels.  Some of these now in the flotilla are dedicated to a high standard of uniform product quality of EVs, produced via safe and fully donor-traceable cell sources like mesenchymal stromal/stem (MSC) cells. [3] [4] When dependable, affordable bioproduction cells and media are incorporated into a validated cGMP process that maps from a 2D culture to fully scalable 3D bioreactor systems, [5] [6] [7] [8] [9] it’s easier for standardized concentration and purification systems to sieve them from the media, free of artifacts or contaminants.  The EV community today responds with a healthy division of labor and interchangeable bioprocess features, spontaneously ordered to contend with the complexity of the exciting new therapies featured at the Conference.  For a clinical developer of EV-based therapeutics, now is a great time to join in the regattaRoosterBio, [10] having itself navigated this space, [11] is here to help simplify the vast complexity and industrialize the EV supply chain for product developers [12] with its own unique cGMP bioprocess expertise and quality source materials with Type II master files, on record with FDA. [13] RoosterBio embraces the complexity and knowledge brought to the EV field by organizations like ISEV and their member scientists.  From our perspective, we say to newcomers, “Come on in, the water’s fine!”


  1. Witcher, Mark. Phase III Clinical Trials – Ever Wonder Why Some Products Unexpectedly Fail? Pharmaceutical Engineering 2019; Available from:
  2. Witcher, Mark. Why Controlling CQAs Isn’t Good Enough For Gene & Cell Therapies. Cell & Gene 2020; Available from:
  3. National Cell Manufacturing, Consortium, Achieving Large-Scale, Cost-Effective, Reproducible Manufacturing of High Quality Cells. A Technology Roadmap to, 2016. 2025: p. 92-101.
  4. RoosterBio. Comparability of hMSC Economic and Quality Attributes after Expansion in Bovine Serum Containing vs Xeno-Free Bioprocessing Media Formulations. 2016; Available from:
  6. Lembong, Josephine, et al., A scalable xeno-free microcarrier suspension bioreactor system for regenerative medicine biomanufacturing of hMSCs. 2019.
  7. Adlerz, Joseph Takacs and Katrina. hUC-MSC Exhibit Robust Proliferation in 3D Bioreactor System. RoosterBio Blog 2020; Available from:
  8. Adlerz, Katrina. Generating MSC-EVs With a Scalable Manufacturing System. RoosterBio Blog 2019; Available from:
  9. Lembong, Josephine. Productivity Metric Considerations in MSC & MSC-EV Manufacturing in Response to COVID-19. RoosterBio Blog 2020; Available from:
  10. Olsen, T. R. and Rowley, J. A., Corporate profile: RoosterBio, Inc. Regen Med, 2018. 13(7): p. 753-757. 10.2217/rme-2018-0092
  11. Ng, K. S., et al., Bioprocess decision support tool for scalable manufacture of extracellular vesicles. Biotechnol Bioeng, 2019. 116(2): p. 307-319. 10.1002/bit.26809
  12. Lembong, J., et al., Bioreactor Parameters for Microcarrier-Based Human MSC Expansion under Xeno-Free Conditions in a Vertical-Wheel System. Bioengineering (Basel), 2020. 7(3). 10.3390/bioengineering7030073
  13. Williams K and Hansen, C. Quality Begins at Inception. RoosterBio Blog 2020; Available from:

RoosterBio is fueling the rapid implementation of scalable advanced therapies. Contact us to discuss how we can accelerate your product & process development. Follow us on LinkedIn for more educational resources just like this.