New test could predict life-threatening side effects of cancer immunotherapy

Medical researchers in Japan have discovered a way to predict a potentially life-threatening side effect of cancer immunotherapy before it occurs. By analyzing cerebrospinal fluid collected pre-treatment, researchers at Kyushu University identified specific proteins associated with a damaging immune response that can affect the central nervous system after therapy. Their findings, published in Leukemia on 11 March, 2025, could make immunotherapy cancer treatment safer by helping doctors identify high-risk patients in advance, allowing early treatment or even prevention of this condition.

Over the past decade, cancer immunotherapy-where the patient's immune system is used to fight cancer-has become a promising new treatment strategy. One form of immunotherapy, called CAR-T-cell therapy, uses genetic engineering to reprogram a patient's immune cells (T cells) to target and destroy cancer cells. It has proven successful in treating blood cancers but comes with serious risks, including immune effector cell-associated neurotoxicity syndrome (ICANS), which causes inflammation in the central nervous system.

ICANS can present with mild symptoms, such as headache or lethargy, but in more severe cases it can be life-threatening, with patients experiencing impaired consciousness, seizures or bleeding in the brain. The incidence rate of ICANS after CAR-T therapy is very high, around 64%, but until now, there hasn't been a reliable way to predict ICANS severity."

Dr. Yuya Kunisaki, Professor, Department of Clinical Chemistry and Laboratory Medicine, Kyushu University Hospital

In this study, the research team analyzed the proteins in leftover cerebrospinal fluid taken from 29 patients with B-cell non-Hodgkin's lymphoma before they received CAR T therapy. Within that cohort, 11 patients developed ICANS while 18 patients did not.

The research team identified 864 proteins present in all spinal fluid samples. From there, they narrowed the list down to 46 proteins that showed clear differences in levels between patients who developed ICANS and those who did not, making them potential biomarkers for predicting the condition.

Ultimately, the researchers found that C1RL, a protein with elevated levels in ICANS patients, and FUCA2, a protein with lower levels in ICANS patients, were the best predictors. When looking at these proteins together, a predictive test determined their ratio had high accuracy in distinguishing patients at high risk of ICANS from those at low risk.

The team then tested the C1RL/FUCA2 biomarker on a second group of 10 patients undergoing CAR-T therapy and found that for all patients, the protein ratio correctly determined the risk of developing ICANS. 

However, the researchers cautioned that despite its high accuracy, the study's small sample size means their findings are still preliminary. "We now need to conduct the study with a larger number of patients to fully validate our results," says co-first author, Dr. Tomoko Nomiyama, a clinical technologist in the Department of Clinical Chemistry and Laboratory Medicine, Kyushu University Hospital.

In addition to helping doctors detect ICANS early and start treatment sooner, the researchers hope that identifying key biomarkers will allow doctors to give preventive drugs before CAR-T therapy begins, reducing the risk of ICANS developing at all. For example, C1RL, which is elevated in ICANS patients, is involved in the complement system-a part of the immune system that is known to cause inflammation and may contribute to ICANS.

"If the biomarker ratio shows a patient is at high risk for ICANS, we could preemptively treat them with drugs that inhibit the complement system to lower the risk," explains Kunisaki. "This predictive test could therefore pave the way for a more personalized and safer approach to cancer treatment."

The research team also plans to test if these biomarkers are accurate for patients with other blood cancers beyond B-cell non-Hodgkin's lymphoma. Finally, they are also widening their search, hoping to uncover key biomarkers in more easily collected fluids, like blood serum.

"Spinal fluid collection is an invasive and painful procedure, so most hospitals in Japan and other countries don't routinely perform it before CAR-T therapy," says Nomiyama. "If we can identify similar biomarkers in blood, our test would become a much simpler and more accessible tool for predicting ICANS."