Knee osteoarthritis affects more than one in ten people over the age of 40, wearing away cartilage and steadily limiting how far and how comfortably they can move. Treatment options are few, and into that gap has stepped a booming market in stem cell injections, some of them offered by clinics selling therapies that have never been properly tested. Even carefully designed clinical trials have failed to settle whether the injections work.

A study published in Stem Cell Reports suggests one reason for the confusion. The cells themselves differ from patient to patient, and those differences may predict who benefits.

Half Responded, Half Did Not

Sowmya Viswanathan of the Schroeder Arthritis Institute at University Health Network in Toronto and her colleagues followed up on a phase I/IIa clinical trial authorized by Health Canada. In it, patients with knee osteoarthritis received injections of mesenchymal stromal cells, or MSCs, harvested from their own bone marrow and grown in the lab.

MSCs, which can be drawn from fat or bone marrow, are widely used in osteoarthritis research. Using a patient’s own cells reduces the risk of an immune reaction and of transmitting infections. The drawback is that the quality and therapeutic potential of those cells vary from one person to the next.

The team tracked 12 patients for 24 months. Anyone who reported at least a 20 percent improvement in both pain and knee mobility counted as a responder. Patients whose pain or mobility stayed the same or worsened counted as non-responders. At 12 months, six of the 12 were responders, and most of those improvements were still present at two years.

The Answer Was in the Cells

To find out what separated the two groups, the researchers went back to the cells. They tested the MSCs from all 12 patients in the laboratory and found clear differences. Cells from responders had a stronger anti-inflammatory effect and were better at suppressing the activity of immune cells.

The team also identified 14 microRNAs, short strands of genetic material that regulate how genes are expressed, present in different amounts in responder and non-responder cells. These could serve as molecular markers of a batch of cells’ therapeutic strength.

What did not predict the outcome was just as telling. Neither the severity of a patient’s arthritis at the start nor their level of inflammation correlated with whether they responded. The difference lay in the cells, not in the knees.

Screening Before Injecting

The practical idea that follows is simple. If laboratory tests on cultured MSCs can reveal their potency in advance, doctors could select only cells with enhanced therapeutic potential for injection, making the treatment more consistent. “Predicting MSC potency could be a powerful tool,” the society’s summary of the work notes.

The study is small, and its authors say larger controlled trials are needed to confirm the findings. With only 12 patients, a 50 percent response rate could shift considerably in a bigger group. Improvement was measured by patients’ own reports of pain and mobility, which are subjective and susceptible to placebo effects, and the summary of the work does not describe a control group that received no cells.

Even so, the work points toward a way out of a field crowded with mixed results. Instead of asking whether stem cell injections work for knee osteoarthritis, it asks which cells work, and whether that can be known before the needle goes in.


Endnotes

  1. Viswanathan, S., et al. (2026). “Responders vs. non-responders to mesenchymal stromal cells in knee osteoarthritis: mechanistic correlates of donor cell and patient features.” Stem Cell Reports. https://doi.org/10.1016/j.stemcr.2026.103024
  2. International Society for Stem Cell Research (August 6, 2026). “New study examines variable outcomes of stem cell injections for knee osteoarthritis and strategies to improve consistency.” EurekAlert! https://www.eurekalert.org/news-releases/1139149

Leave a Reply

Trending