Last week’s much anticipated FDA approval of the first chimeric antigen receptor T-cell (CAR-T) therapy for acute lymphocytic leukemia hails as the first gene therapy on the US  market.

Classified as a “cell-based gene therapy,” Novartis’ (Basel, Switzerland) Kymriah works by removing patients’ T-cells, using a viral vector to introduce a gene that will allow the T-cells to recognize and kill cancer, and then infusing these modified T-cells back into the patient. Recall T-cells are found in the blood and fight disease.

Along with its significant potential, Kymriah also carries serious risks. Its approval came with a boxed warning because of the potential for “cytokine release syndrome (CRS),” also referred to as a cytokine storm, which has caused fatalities in clinical trials of other CAR-T products.

Making CAR-T safer while maintaining efficacy are goals of next generation CAR-T. Let’s explore cytokine storms and find out how scientists aim to circumvent this roadblock to fighting cancer.


Cytokines are small proteins which play an important role in relaying messages from one cell to surrounding cells and tissue. Cytokines serve two main functions involving white blood cells:

  • Activate additional white blood cells to fight off pathogens
  • Stimulate white blood cells to move towards sites of inflammation

Cytokine signaling makes for a very quick and strong immune response. Usually, the response is kept in check, and dissipates when the bad cells have been eliminated.

However, in some cases, this positive feedback loop — activated cells releasing still more activating cytokines — spins out of control, resulting in a cytokine storm. Acute inflammation with accompanying symptoms such as high fever, swelling, and nausea can occur. In severe cases, serious tissue damage and death can result — for example, lung failure induced by excessive amounts of fluids and cells moving into the lungs.

A cytokine storm is the adverse event most associated with CAR-T treatments and next generation CAR-T treatments are being developed to have built-in controls to regulate cytokines so the storms can be stopped.


In first generation CAR-T, maximum activation occurs. With a full cytokine barrage, there is no way to tamp down the cytokine response.

In next generation CAR-T, a small molecule drug may be co-administrated with the therapy. The drug’s function is to activate CAR-T to fight cancer, or turn it “on.” If a cytokine storm ensues, the small molecule drug can be immediately withdrawn — the “off” switch — essentially deactivating CAR-T and stopping cytokine release.

This second iteration of CAR-T is made possible by a handful of companies who are designing drugs that will act as an “on/off switch” to control CAR-T. Bellicum Pharmaceuticals (Houston, TX) is developing a CAR-T product, BPX-601, that uses small molecule-activation. BPX-601 entered Phase I clinical trials in February 2017. Intrexon (Germantown, MD) has a similar product in Phase I clinical development.


Juno Therapeutics (Seattle, WA) is developing a CAR-T product that uses a sort of two-step verification process. It turns out that tumor cells have many proteins on their surface — so it is a challenge to find a distinct protein that is also unique to any given cancer cell. Instead of relying upon finding that one special protein, why not target a more common one and use another protein to double check the work?

Juno’s approach: bispecific chimeric antigen receptors. This means each engineered killer T-cell has not one, but two chimeric antigen receptors (CARs).

  • One CAR is activated in the presence of a protein found on the surface of cancer cells. Once activated this CAR-T cell would produce more copies of itself, release cytokines, and attack the tumor.
  • A second CAR called an inhibitory CAR (iCAR), is activated in the presence of a different protein found only on healthy cells — NOT on cancer cells. If an iCAR is activated, an inhibitory signal is sent to the first CAR, preventing the CAR-T from working.

Simply put, one CAR finds the target protein while the other iCAR verifies the cell is “unhealthy” via another protein. This bispecific CAR-T, now in preclinical development, aims to eliminate the off-target effects and decrease the amount of cytokines released.


Earlier this year, Cellectis (Paris, France) published a paper describing work they’ve done to engineer CAR-Ts with an oxygen-sensitive domain.

Under normal cellular conditions, this domain signals the CAR-T to remain inactive. Under low oxygen, or hypoxic conditions, this domain sends an activation signal to the CAR-T. Since most solid tumors have a hypoxic environment, an oxygen-sensitive CAR-T should be activated within the tumor but not outside of it.


It turns out that CAR-T activation results in cytokine production because part of the CAR-T is actually inside the cancer cell, where it interacts with other proteins in a cytokine signaling pathway.

Scientists at the Blood Research Institute of Blood Center of Wisconsin (Milwaukee, WI) have developed decoy molecules that interfere with the protein-protein interactions in the pathway. These decoy molecules are short protein fragments called peptides. By binding to proteins in the cytokine pathway, the signal to produce more cytokines is blocked.

The decoy peptides reduce cytokine production by ~70%, which is likely enough to prevent the immune response from spinning out of control. In the words of Laura Savatski of the Blood Research Institute, “By reducing cytokine production by CAR-T cells, you prevent a cytokine storm from happening. So instead of dealing with a problem at the back end of the therapy, you solve it at the front end through intelligent design.” Decoy molecules are currently in preclinical development.

With some of the best minds in the biopharma industry working on CAR-T design, this landmark FDA approval is likely to be just the first shot in a treatment revolution.


Share This