How Is Stem Cell Therapy Done, Step by Step?

Stem cell therapy is not one procedure but a chain of tightly coordinated steps, and the specifics shift depending on the type of cells used, the disease being treated, and whether the cells come from the patient or a donor. At its broadest, the sequence moves from harvesting cells, to isolating and expanding them in a laboratory, to delivering them into the body and monitoring what happens next. Each of those stages carries its own technical challenges, and the gap between a bone marrow transplant for leukemia and an experimental injection of fat-derived stem cells for a knee injury is enormous. Understanding what actually happens at each step makes it easier to evaluate whether a particular therapy is grounded in evidence or riding on hype.

Where the Cells Come From

The starting material for stem cell therapy can be pulled from several places in the body, and the source matters more than most people realize. Bone marrow has been the classic reservoir since the first transplants decades ago. A needle is inserted into the back of the pelvic bone under anesthesia, and liquid marrow is drawn out in a procedure called aspiration. The marrow contains hematopoietic stem cells, which give rise to all blood cell types, along with mesenchymal stem cells that can become bone, cartilage, and fat tissue.

Adipose (fat) tissue is the other major source for mesenchymal stem cells. Fat is collected through a small liposuction procedure, then processed in the lab to break down the tissue and release the cells embedded within it. One large comparative study evaluating over 180 donors found that adipose-derived stem cells grew more reliably in culture than bone marrow stem cells regardless of patient age, sex, or body mass index, and they also produced higher levels of a key bone-repair protein after genetic modification.¹ That difference in growth characteristics is one reason adipose tissue has gained popularity for orthopedic and regenerative applications.

The lab processing of fat tissue itself involves washing the sample, chopping or digesting it with an enzyme called collagenase to free the cells, filtering out debris, and spinning the mixture in a centrifuge to collect a cell pellet at the bottom of the tube.² That pellet, sometimes called the stromal vascular fraction, is the raw starting material for everything that follows.

For blood cancers and certain genetic diseases, the cells often come from peripheral blood rather than marrow. The patient or donor receives injections of growth factors, most commonly G-CSF, over several days. These drugs coax stem cells out of the bone marrow and into the bloodstream, a process called mobilization. Newer mobilization agents work by blocking a receptor called CXCR4 that normally anchors stem cells in the marrow. Studies of G-CSF combined with one such agent (plerixafor) have shown that more stem cells appear in the bloodstream when that anchoring signal is disrupted.³ Once enough stem cells are circulating, blood is drawn through an IV line, run through a machine that skims off the stem cell layer, and the remaining blood is returned to the body. This is called apheresis, and it typically takes a few hours.

Isolating and Purifying the Right Cells

A bone marrow aspirate or a bag of apheresis product is a messy mixture of many cell types. The next step is separating out the stem cells you actually want. The workhorse technique is density gradient centrifugation: the sample is layered over a dense liquid and spun, causing different cell types to settle into distinct bands based on their weight. Researchers have worked to improve this step using specialized tubes that make the separation faster and more reproducible.⁴

For applications requiring very high purity, labs use magnetic bead sorting. Tiny magnetic particles are coated with antibodies that stick to a specific marker on the stem cell surface. The sample is passed through a magnetic column, which holds back the labeled cells while everything else washes through. This approach routinely achieves purities above 98 percent for hematopoietic stem cells.⁵ The same principle works for skin stem cells and reproductive stem cells by swapping in a different antibody.

Growing and Preparing the Cells in the Lab

Often the number of stem cells collected from a patient or donor is not enough for treatment. The cells need to be expanded, which means growing them in culture dishes or bioreactors under carefully controlled conditions. For mesenchymal stem cells, this usually involves seeding the cells onto a surface, feeding them nutrient-rich growth medium, and letting them multiply over days to weeks. The challenge is scaling up without losing the cells’ therapeutic properties.

Bioreactor technology has made large-scale expansion feasible. One approach uses microcarrier beads suspended in a spinning vessel, giving cells much more surface area than a flat dish. Researchers have achieved cell densities above one million cells per milliliter using this method while keeping the cells in their undifferentiated state and maintaining a normal chromosome count.⁶ Running these cultures without animal-derived ingredients (called xeno-free conditions) is increasingly standard, since animal proteins can trigger immune reactions in patients.

Expansion is not without risks. Prolonged culture can push stem cells toward a state resembling cellular aging, where the cells flatten out, lose their characteristic surface markers, and begin secreting stress signals.⁷ Even more concerning, one study found that adipose-derived stem cells accumulated progressive DNA damage and increased chromosomal abnormalities during bioreactor expansion, despite maintaining high viability and normal surface markers throughout the process.⁸ This means cells can look healthy by standard measures while silently accumulating genetic damage, which is why quality control testing at the end of expansion is so critical.

When the goal is not just more cells but a specific cell type, the lab adds another stage called directed differentiation. By exposing stem cells to defined sequences of growth factors at precise concentrations, researchers can steer them toward becoming liver cells, kidney cells, thyroid cells, or other specialized tissues.⁹ One published protocol generates mature kidney podocytes from induced pluripotent stem cells with over 90 percent efficiency within 26 days.¹⁰ These protocols are essentially recipes that mimic the chemical signals a developing embryo uses to build organs, compressed into weeks on a lab bench.

Quality Control Before the Cells Leave the Lab

Before any cell product is administered to a patient in a regulated setting, it has to pass a battery of release tests. For mesenchymal stem cells, this means confirming the cells display the right surface markers in the right proportions and that they retain the ability to differentiate into bone, fat, and cartilage lineages.¹¹ Labs also screen for sterility, checking that no bacteria, fungi, or viruses have contaminated the culture.

For induced pluripotent stem cells, quality control is even more involved. One validated testing framework requires that at least three markers of the undifferentiated state be expressed on at least 75 percent of the cells, alongside checks for residual genetic material left over from the reprogramming process and confirmation that the cells can still differentiate on command.¹² These tests exist because a batch that looks fine under a microscope might harbor undifferentiated cells capable of forming tumors, or might have lost its therapeutic potential entirely.

Freezing and Banking

Many stem cell products are cryopreserved, meaning frozen in liquid nitrogen, either for storage or for shipping to the treatment site. Freezing introduces its own set of problems. Research on bone marrow mesenchymal stem cells found that thawed cells had lower viability, higher rates of early cell death, and reduced ability to stick to surfaces during the first four hours after thawing. Even 24 hours post-thaw, metabolic activity and adhesion had not fully recovered.¹³ The practical takeaway is that how and when cells are thawed before administration matters, and some clinics now build in a recovery period after thawing before infusing cells into a patient.

Conditioning the Patient

For hematopoietic stem cell transplants used to treat blood cancers, there is a critical step before the cells are infused: conditioning. This involves chemotherapy, radiation, or both, given at high enough doses to wipe out the patient’s diseased bone marrow and suppress the immune system enough to accept the incoming cells.¹⁴ The conditioning regimen is tailored to the disease, the patient’s age, and their overall health. Older or frailer patients typically receive reduced-intensity conditioning, which uses lower doses and relies more on the incoming immune cells to clear the remaining disease. Younger patients with aggressive cancers may get myeloablative conditioning, essentially the highest tolerable doses. Some protocols combine a chemotherapy drug with total body irradiation for patients at high risk of relapse.¹⁵

For non-transplant stem cell therapies, like injecting mesenchymal stem cells into a knee joint or the bloodstream for an autoimmune condition, conditioning is usually not part of the process at all. The patient may simply show up, receive the injection, and go home.

Delivering the Cells

The route of delivery depends entirely on the target. Hematopoietic stem cells for blood disorders are almost always infused intravenously, essentially given like a blood transfusion. The cells travel through the bloodstream and, if all goes well, find their way to the bone marrow and begin producing new blood cells. A chemical signal called SDF-1 (also known as CXCL12), produced by cells in the bone marrow niche, acts as a homing beacon for transplanted stem cells expressing the matching receptor, CXCR4.¹⁶

Intravenous delivery has a well-known limitation: many of the infused cells get trapped in the lungs, liver, and spleen before they ever reach the intended tissue. A primate study comparing intravenous to intraosseous (directly into the bone) injection found that intravenous delivery trapped two to five times more donor cells in peripheral organs, while direct bone injection resulted in about a sixfold increase in cells retained at the injection site.¹⁷ For orthopedic applications, cells are often injected directly into a joint or a fracture site. For heart repair, they may be delivered via a catheter threaded into a coronary artery or injected directly into the heart muscle during surgery.

A growing area of research involves embedding stem cells within scaffolds or hydrogels before delivery. These materials mimic the body’s own structural matrix and can hold cells in place at the repair site rather than letting them drift away. Hydrogels designed to match the architecture of natural tissue are being tested for bone repair and cardiac regeneration, offering more precise control over where the cells end up and how they behave after implantation.¹⁸¹⁹

How the Cells Actually Work Once Inside You

Early on, the hope was that transplanted stem cells would physically replace damaged tissue by engrafting and turning into new heart cells, cartilage, or neurons. That picture has largely given way to something more nuanced. For mesenchymal stem cells in particular, engraftment rates in vivo are low and transient. The therapeutic benefit instead comes mainly from what the cells secrete: a cocktail of growth factors, anti-inflammatory signals, and tiny membrane-bound packages called extracellular vesicles.²⁰ This paracrine effect can protect existing tissue from further damage, recruit blood vessel growth, and tamp down harmful inflammation.²¹

The discovery that the secreted factors, rather than the cells themselves, do much of the heavy lifting has opened an entirely new line of research. Some groups are now investigating whether you can skip the cells altogether and just deliver the secretome, the collection of proteins and vesicles the cells produce, as a cell-free therapy.²² If that works, it would sidestep many of the manufacturing, safety, and storage challenges that come with handling living cells.

The local environment also shapes how well the cells perform. Modeling studies have shown that injured or oxygen-deprived tissue releases much higher levels of SDF-1, dramatically increasing the number of stem cells that home to the damaged area. Pre-treating cells with low oxygen before transplantation more than doubled their active homing to injured liver tissue in animal models and improved their survival after arrival.²³ Experimental techniques like ultrasonic microbubble treatment can also boost expression of the homing receptor on the cell surface, improving the rate at which the cells reach their target.²⁴

Tracking the Cells After Delivery

Once cells have been infused or injected, clinicians need to know whether they arrived at the right place and whether they survived. Several imaging methods are used for this in clinical and preclinical settings. MRI is common because it offers high-resolution anatomical images without radiation; cells are labeled with iron oxide nanoparticles before transplantation so they show up as dark spots on the scan. PET imaging uses radioactive tracers to provide functional data about the cells’ activity and can track them over time.²⁵ Optical imaging methods, including bioluminescence and fluorescence, are widely used in animal research but are limited in humans because light does not penetrate deep tissue well.²⁶

No single imaging technique is perfect. MRI has excellent spatial detail but cannot easily distinguish live cells from dead ones that still contain the iron label. PET can detect very small numbers of cells but exposes the patient to radiation and has lower resolution. Multimodal approaches, combining two techniques to get both structural and functional information, are an active area of development.²⁷

Safety and the Risk of Unproven Treatments

The safety profile of stem cell therapy varies enormously depending on what kind of cells are used and how the procedure is done. Hematopoietic stem cell transplants for cancer are a mature medical procedure with well-characterized risks: infection during the period when the immune system is suppressed, graft-versus-host disease when donor immune cells attack the recipient’s body, and organ toxicity from the conditioning regimen. In one trial comparing autologous and allogeneic transplants for T-cell lymphoma, about a third of patients who received a donor transplant died from transplant-related toxicity, while none died of toxicity after receiving their own cells back.²⁸ That trade-off between disease control and treatment-related death is at the core of every transplant decision.

For newer, less established therapies, the concern is different. Induced pluripotent stem cells carry a risk of uncontrolled growth or undesired differentiation, where cells meant to become one tissue type instead become another, or form tumors called teratomas.²⁹ Mesenchymal stem cells are generally considered safer, but laboratory studies have raised concerns about their potential to promote tumor growth and metastasis in certain contexts.

Then there is the sprawling market of clinics selling stem cell treatments that have never been through rigorous testing. A detailed 2021 survey documented hundreds of U.S. businesses selling unlicensed stem cell products for conditions ranging from Alzheimer’s disease to erectile dysfunction, with published case reports of patients suffering serious harm including blindness and infections.³⁰ These clinics typically skip most of the steps described above. They might harvest fat, minimally process it, and reinject it the same day without proper isolation, expansion, quality control, or any evidence that the product does what they claim. The FDA considers many of these products to be unapproved biological drugs, and adverse events from them are likely underreported.

Manufacturing at Scale

One of the largest obstacles to making stem cell therapies widely available is manufacturing. Producing a personalized cell product for one patient at a time is extraordinarily expensive, and the process is difficult to standardize. Existing methods for autologous therapies face challenges related to cost, variability in the starting material, safety risks during handling, and scalability.⁷ Automated, closed-system bioreactors are being developed to reduce human error and contamination risk, and artificial intelligence is being explored to optimize culture conditions in real time.

The field is also moving toward off-the-shelf allogeneic products, where cells from a single healthy donor are expanded, banked, and used to treat many patients. This would dramatically cut costs and turnaround times, but it introduces immunological challenges since the recipient’s body may reject foreign cells. How well cryopreservation works for these banked products, and how quickly cells recover their function after thawing, become central questions for this model. The shift from bespoke, patient-by-patient manufacturing toward something closer to industrial production is probably the single biggest factor that will determine whether stem cell therapies move from specialized centers to routine medicine.