George F. Muschler, MD, FAAOS, FAOA, FORS, FAAAS


Published 12/20/2023
Terry Stanton

Finding the Right Cell: Using Automation to Select Effective Biologic Therapies

The third annual AAOS Biologics Symposium, titled “The Next Generation of Biologics,” which took place on July 12 in Washington, D.C., featured four sessions on regulatory clarity, autologous tissue processing, personalized blood products, and future biologics trends. At the event, George F. Muschler, MD, FAAOS, FAOA, FORS, FAAAS, professor of orthopaedics and biomedical engineering and director of the Joint Preservation Center at Cleveland Clinic, presented the session “Automated Methods for Characterization of Cell Therapy Formulations.” In a Q&A with AAOS Now, Dr. Muschler elaborated on advances in selecting cells that have therapeutic utility and new technology for performance-based selection.

AAOS Now: A central theme of your presentation related to the move toward data-driven decision making and especially the quantification of biologics. Could you expand on this concept?
Dr. Muschler: Musculoskeletal medicine has benefited greatly from several waves of scientific innovation, from biomechanics, metallurgy, polymers, tribology, scaffold materials, clinical imaging, osteotropic drugs, to arthroscopic imaging and instrumentation. Clinical use of cells and cell-derived bioactive biomaterials may be poised to provide the next wave of new therapies and adjuvants, opening the door to disease-modifying and regenerative therapies.

The feasibility of cellular therapies using autologous or even engineered allograft cell populations is clear. However, it is important for each of us to remember that this is not a time for hype or overconfidence. There is a rigorous path ahead of us: to define critical quality attributes that are needed for these cells or cell products; build highly repeatable and reproducible methods for fabrication; design rigorous quantitative metrics that strongly predict performance; define means to control the dose and composition of cells being delivered; develop transport, storage, and delivery systems; and create means to understand, monitor, and mitigate risks, such as infection or neoplasm. This requires engagement across a broad range of expertise and a commitment to regulatory standards, clinical registries, and clinical trial networks. These are critical next steps to providing us with the data needed to ensure that our patients are offered safe and effective therapies and to refine [clinical practice guidelines], [appropriate use criteria], and appropriate payment models.

What is the typical approach to culturing and expanding cells for processing? 
There are three key steps in cell processing: cell sourcing (i.e., starting with the right cell[s]); control of the environment where cells are processed or cultures are expanded; and precise characterization of the endpoint (i.e., confirming an effective and reproducible product).

Currently, we generally source cells from bone marrow aspirates or emulsified fat. These sources have the advantages of a very low cost and morbidity. However, they also have a disadvantage in that every patient and sample contains a different population of cells, and few of these cells are progenitors. On average, one in 2,000 fat cells and one in 200,000 cells from bone marrow aspirates are a connective tissue progenitor (CTP), or a cell that is capable of proliferation to form progeny and differentiate into one or more connective tissues (e.g., bone, cartilage, fat). But the prevalence of these cells in any individual or tissue can vary by 100-fold or more. Fortunately, these progenitors will rapidly attach to tissue culture plastic and then divide rapidly under known conditions to generate millions of cells, but we don’t always get what we want at the end of this expansion.

At first glance, this is a simple process to execute. However, each tissue sample contains a mixture of many cell types and CTPs, but not all have the biological potential that we seek. Growing cells on a dish results in a process of profound selection. We eliminate blood-forming cells.

Traditionally, we have allowed many CTPs to grow together and compete with one another for space and nutrients. As the entire population expands, this selects out the most rapidly growing clones, which dilute out the contribution of slower-growing (but perhaps more desirable) clones. Over 2 to 5 weeks, 50 to 100 original CTPs can expand to become 6 million to 10 million cells. These culture-expanded polyclonal populations can be referred to as mesenchymal stromal cells (MSCs) if they meet certain criteria set by the International Society of Cell and Gene Therapy.

When we culture cells at low density, so that individual progenitors are separated, each CTP will form its own patch of progeny or daughter cells (a colony). The number of colonies that form can be used to estimate the number of CTPs in the starting population. Moreover, each of these colonies is different, and the differences between the original colony-founding CTP are disclosed by the differences in proliferation rate, cell migration, and differentiation features (morphology and gene expression).

Knowing that CTPs differ from one another has inspired great interest in selection of individual clones or clone types as the starting material for cell products, so that the complete provenance of the cells that are going into a patient is known. Selection of some progenitors over others will, in theory, provide greater control over quality and reproducibility of the final cell product and potentially allow adaptation to the deficiencies in progenitor populations that result from aging or disease.

There is an additional refocusing taking place in manufacturing, which involves improving our understanding of the process of in vitro competition and changing culture conditions so the most therapeutic cells have the greatest competitive advantage and will be preferentially expanded and preserved. Characterization of the cells that come out at the end is inevitably an important part of the process of delivering a therapy. Both patients and the FDA require products that are safe and effective and whose identity, dose, and critical quality attributes are known to provide an optimal therapeutic benefit.

Where are we in the overall process?
We need alternatives for patients who can’t be treated effectively with current therapies. However, those alternatives need to be safe, effective, repeatable, reproducible, and approved through the FDA and other regulatory bodies. So far, there is increasing evidence that it’s possible to grow some cells in culture under defined conditions and then transplant them into patients with a low risk of serious adverse effects, if certain rules are followed.

An effective therapy requires much more than safety. It requires that the right cell is placed in the right location in the right way at the right dose for the right reason. Experience from several trials has now shown us that culture-expanded MSC populations can be placed in certain locations and at certain doses without causing harm. The challenge now is to go to phase 2 and phase 3 trials in specific indications to test the question of efficacy.

It is also important to recognize that we may not have the right cells yet. Cells in future trials may not match current criteria for MSCs. We may need to enhance or modify the biological performance of cells in specific ways to provide value in clinical settings. There is increasing interest in modifying cells to improve their performance, either improving their survival, reducing immunogenicity, driving preferred differentiation, or modulating secretion of certain therapeutic chemicals. Some modifications can be accomplished transiently using some types of viral vectors; others must be engineered as permanent changes to a cell population that is preserved in all the future progeny.

What laboratory tools are available for the process of performance-based clone selection for cell fabrication?
There are several evolving platforms designed to enable automated cell processing but only a few designed to enable specific clone selection. Our lab has been working to develop one such platform using a robot that is built around a fluorescence microscope. This allows a user growing cells in a cell culture dish to (1) collect high-resolution images to “see all the cells in their garden,” (2) analyze those images and make decisions about which cells they want to keep and which cells they want to get rid of, and then (3) to “pick” or “weed out” specific colonies or cells—all using the robot. The value of this approach is not only increased precision but also exceptional documentation of every step in the process. It’s a platform that is finding traction, not only in the expansion of MSCs but also in the generation of induced pluripotent cells that can be made into an MSC-like population or directly made into cartilage cells, nerve cells, etc., in vitro.

Do you anticipate that machine learning will contribute to the selection of which cells are expanded?
Artificial intelligence and machine-learning methods will have a profound effect on the development of cell therapies and products. These technologies are already enhancing: (1) the information that we can extract from individual images of cells and colonies, (2) analysis of cell state and process variables that predict outcome, and (3) multidimensional process optimization. Users will generate roadmaps of what success looks like very early in the process and have an early warning system when cell features begin to fall off track.

Any other takeaways you want to address?
I think the field is really in a threshold where there are a lot of exciting things happening. It’s easy to become excited, and in some ways, it is easy to get overexcited because none of this is going to happen tomorrow or the day after in terms of effective therapies. On the other hand, it’s also possible to be overly pessimistic, to look at the complexity that we still have to solve and control and to document in order to have safe, effective therapies. The truth is that the tools that we need are coming online quickly. We are increasingly able to grapple with complexity that has been daunting in the past, and I feel confident that reproducible fabrication of high-quality cells will become routinely manageable in the not-too-distant future. I encourage all of us to participate in high-quality clinical trials and to be vigilant, patient, informed, and objective as we keep our patients informed about these evolving opportunities.

Terry Stanton is the senior medical writer for AAOS Now. He can be reached at

Read more about the AAOS Biologics Symposium

Read the AAOS Now article “AAOS Biologics Symposium Investigates the Present and Future of Orthobiologic Therapies” for a summary of the AAOS Biologics Symposium.