A new platform designed to forecast ‘cytokine storms’

The Jackson Laboratory has developed a new platform to provide a fast, accurate and reliable method for screening novel immunotherapy treatments for efficacy and cytokine release syndrome.

Immunotherapies have become powerful tools for treating serious diseases, from cancer to immune disorders. But in some patients, these drugs can trigger cytokine release syndrome, a condition with symptoms ranging from mild fever to life-threatening organ failure.

Until now, drug screening methods have failed to reliably predict which experimental treatments are more likely to cause these ‘cytokine storms’. This new service, offered exclusively from The Jackson Laboratory, can predict outcomes before treatments make it to the clinic. Additionally, this could also be useful for potentially predicting which patients are more likely to develop cytokine storms in response to certain drugs.

“One of the problems with immunotherapies is that you don’t know how different people are going to react to the same therapy,” says Dr. James Keck, Senior Director of Innovation and Product Development at The Jackson Laboratory.

Several monoclonal antibody therapies have been linked to cytokine release syndrome. In 2006, a human trial on the leukemia drug theralizumab (TGN1412) resulted in high cytokine levels and organ failure in healthy volunteers, despite preclinical studies finding no evidence of cytokine storm potential.

To develop the model, Keck and his team used three different immunodeficient mouse strains engrafted with the peripheral blood mononuclear cells (PBMC) from 10 healthy human donors. These recipients were then used to compare cytokine responses to three different antibody-based therapies known to stimulate cytokine release, including muromonab-CD3 (OKT3), anti-CD28 and an analog of TGN1412.

When the researchers treated the mice with OKT3 or another monoclonal antibody, anti-CD28, they found that human cytokine levels varied across the 10 donors. For example, the PBMC humanized mice treated with OKT3 resulted in a significant spike in the human cytokine interferon gamma for eight donors, while only three treated with anti-CD28 showed elevated levels. They also observed the response to treatment was highly reproducible when repeated with a given human PBMC donor.

“Even though all the donors were healthy and normal, they all reacted differently to the same therapy,” says Keck, who led the study. “That’s very important to know individual variability because that variability is often seen in the clinic.”

To investigate whether the new method was more sensitive than current approaches, the team compared cytokine levels in their new in vivo model with cytokine levels in cell cultures after treatment with OKT3 or anti-CDT28 using five donors. For both treatments, human cytokine levels were higher and more consistent in the mice than the cultures. The culture assay revealed both false negatives and false positives, indicating that the in vivo engraftment model can pinpoint responders that would go undetected by standard culture screening methods.

The researchers also used a mouse strain that expresses the human stem cell factor GM-CSF, IL3, and Kit ligand to assess whether the model could be used to predict cytokine release syndrome in response to a TGN1412 analogue. The TGN1412 CD28 superagonist was tested at three doses using cells from the same donor and showed a clear increase in response associated with increased dosing, validating that the model could be used as an accurate prediction tool.

In addition to screening immunotherapy drugs for potential clinical trials, Keck says that the platform and service could also help researchers evaluate drugs for inhibiting cytokine release syndrome in patients with severe COVID-19.

“Many of the same cytokines induced in our models are also seen in COVID-19 patients,” Keck says. “We are using the model to help researchers screen for drugs to inhibit cytokine release syndrome, which will hopefully find their way one day into the clinic.”