Predicting neurotoxicity after CAR T-cell therapy

Fred Hutch News Service

Researchers say they have identified potential biomarkers that may be used to help identify patients at an increased risk of neurotoxicity after chimeric antigen receptor (CAR) T-cell therapy.

The team also created an algorithm intended to identify patients whose symptoms were most likely to be life-threatening.

The researchers discovered the biomarkers and developed the algorithm based on data from a trial of JCAR014, an anti-CD19 CAR T-cell therapy, in patients with B-cell malignancies.

Cameron J. Turtle, MBBS, PhD, of Fred Hutchinson Cancer Research Center in Seattle, Washington, and his colleagues described this research in Cancer Discovery.

“It’s essential that we understand the potential side effects of CAR T therapies” Dr Turtle said. “While use of these cell therapies is likely to dramatically increase because they’ve been so effective in patients with resistant or refractory B-cell malignancies, there is still much to learn.”

Dr Turtle and his colleagues sought to provide a detailed clinical, radiological, and pathological characterization of neurotoxicity arising from anti-CD19 CAR T-cell therapy.

So the team analyzed data from a phase 1/2 trial of 133 adults with relapsed and/or refractory CD19+ B-cell acute lymphoblastic leukemia, non-Hodgkin lymphoma, or chronic lymphocytic leukemia.

The patients received lymphodepleting chemotherapy followed by an infusion of JCAR014.

Neurotoxicity

Within 28 days of treatment, 53 patients (40%) developed grade 1 or higher neurologic adverse events (AEs), 28 patients (21%) had grade 3 or higher neurotoxicity, and 4 patients (3%) developed fatal neurotoxicity.

Of the 53 patients with any neurologic AE, 48 (91%) also had cytokine release syndrome (CRS). All neurologic AEs in the 5 patients who did not have CRS were mild (grade 1) and transient.

Neurologic AEs included delirium with preserved alertness (66%), headache (55%), language disturbance (34%), decreased level of consciousness (25%), seizures (8%), and macroscopic intracranial hemorrhage (2%).

For most patients, neurotoxicity resolved by day 28 after CAR T-cell infusion. The exceptions were 1 patient in whom a grade 1 neurologic AE resolved 2 months after CAR T-cell infusion and the 4 patients who died of neurotoxicity.

The 4 neurotoxicity-related deaths were due to:

• Acute cerebral edema (n=2)
• Multifocal brainstem hemorrhage and edema associated with disseminated intravascular coagulation (n=1)
• Cortical laminar necrosis with a persistent minimally conscious state until death (n=1).

Potential biomarkers

In a univariate analysis, neurotoxicity was significantly more frequent in patients who:

• Had CRS (P<0.0001)
• Received a high CAR T-cell dose (P<0.0001)
• Had pre-existing neurologic comorbidities at baseline (P=0.0059).

In a multivariable analysis (which did not include CRS as a variable), patients had an increased risk of neurotoxicity if they:

• Had pre-existing neurologic comorbidities (P=0.0023)
• Received cyclophosphamide and fludarabine lymphodepletion (P=0.0259)
• Received a higher CAR T-cell dose (P=0.0009)
• Had a higher burden of malignant CD19+ B cells in the bone marrow (P=0.0165).

The researchers noted that patients who developed grade 3 or higher neurotoxicity had more severe CRS (P<0.0001).

“It appears that cytokine release syndrome is probably necessary for most cases of severe neurotoxicity, but, in terms of what triggers a person with cytokine release syndrome to get neurotoxicity, that’s something we need to investigate further,” said study author Kevin Hay, MD, of Fred Hutchinson Cancer Research Center.

Dr Hay and his colleagues also found that patients with severe neurotoxicity exhibited evidence of endothelial activation, which could contribute to manifestations such as capillary leak, disseminated intravascular coagulation, and disruption of the blood-brain barrier.

Algorithm

The researchers developed a predictive classification tree algorithm to identify patients who have an increased risk of severe neurotoxicity.

The algorithm suggests patients who meet the following criteria in the first 36 hours after CAR T-cell infusion have a high risk of grade 4-5 neurotoxicity:

• Fever of 38.9°C or greater
• Serum levels of IL6 at 16 pg/mL or higher
• Serum levels of MCP1 at 1343.5 pg/mL or higher.

This algorithm predicted severe neurotoxicity with 100% sensitivity and 94% specificity. Eight patients were misclassified, 1 of whom did not subsequently develop grade 2-3 neurotoxicity and/or grade 2 or higher CRS.

Funding

This research was funded by Juno Therapeutics Inc. (the company developing JCAR014), the National Cancer Institute, Life Science Discovery Fund, the Bezos family, the University of British Columbia Clinical Investigator Program, and via institutional funds from Bloodworks Northwest.

Dr Turtle receives research funding from Juno Therapeutics, holds patents licensed by Juno, and has pending patent applications that could be licensed by nonprofit institutions and for-profit companies, including Juno.

The Fred Hutchinson Cancer Research Center has a financial interest in Juno and receives licensing and other payments from the company.

Fred Hutch News Service

Researchers say they have identified potential biomarkers that may be used to help identify patients at an increased risk of neurotoxicity after chimeric antigen receptor (CAR) T-cell therapy.

The team also created an algorithm intended to identify patients whose symptoms were most likely to be life-threatening.

The researchers discovered the biomarkers and developed the algorithm based on data from a trial of JCAR014, an anti-CD19 CAR T-cell therapy, in patients with B-cell malignancies.

Cameron J. Turtle, MBBS, PhD, of Fred Hutchinson Cancer Research Center in Seattle, Washington, and his colleagues described this research in Cancer Discovery.

“It’s essential that we understand the potential side effects of CAR T therapies” Dr Turtle said. “While use of these cell therapies is likely to dramatically increase because they’ve been so effective in patients with resistant or refractory B-cell malignancies, there is still much to learn.”

Dr Turtle and his colleagues sought to provide a detailed clinical, radiological, and pathological characterization of neurotoxicity arising from anti-CD19 CAR T-cell therapy.

So the team analyzed data from a phase 1/2 trial of 133 adults with relapsed and/or refractory CD19+ B-cell acute lymphoblastic leukemia, non-Hodgkin lymphoma, or chronic lymphocytic leukemia.

The patients received lymphodepleting chemotherapy followed by an infusion of JCAR014.

Neurotoxicity

Within 28 days of treatment, 53 patients (40%) developed grade 1 or higher neurologic adverse events (AEs), 28 patients (21%) had grade 3 or higher neurotoxicity, and 4 patients (3%) developed fatal neurotoxicity.

Of the 53 patients with any neurologic AE, 48 (91%) also had cytokine release syndrome (CRS). All neurologic AEs in the 5 patients who did not have CRS were mild (grade 1) and transient.

Neurologic AEs included delirium with preserved alertness (66%), headache (55%), language disturbance (34%), decreased level of consciousness (25%), seizures (8%), and macroscopic intracranial hemorrhage (2%).

For most patients, neurotoxicity resolved by day 28 after CAR T-cell infusion. The exceptions were 1 patient in whom a grade 1 neurologic AE resolved 2 months after CAR T-cell infusion and the 4 patients who died of neurotoxicity.

The 4 neurotoxicity-related deaths were due to:

• Acute cerebral edema (n=2)
• Multifocal brainstem hemorrhage and edema associated with disseminated intravascular coagulation (n=1)
• Cortical laminar necrosis with a persistent minimally conscious state until death (n=1).

Potential biomarkers

In a univariate analysis, neurotoxicity was significantly more frequent in patients who:

• Had CRS (P<0.0001)
• Received a high CAR T-cell dose (P<0.0001)
• Had pre-existing neurologic comorbidities at baseline (P=0.0059).

In a multivariable analysis (which did not include CRS as a variable), patients had an increased risk of neurotoxicity if they:

• Had pre-existing neurologic comorbidities (P=0.0023)
• Received cyclophosphamide and fludarabine lymphodepletion (P=0.0259)
• Received a higher CAR T-cell dose (P=0.0009)
• Had a higher burden of malignant CD19+ B cells in the bone marrow (P=0.0165).

The researchers noted that patients who developed grade 3 or higher neurotoxicity had more severe CRS (P<0.0001).

“It appears that cytokine release syndrome is probably necessary for most cases of severe neurotoxicity, but, in terms of what triggers a person with cytokine release syndrome to get neurotoxicity, that’s something we need to investigate further,” said study author Kevin Hay, MD, of Fred Hutchinson Cancer Research Center.

Dr Hay and his colleagues also found that patients with severe neurotoxicity exhibited evidence of endothelial activation, which could contribute to manifestations such as capillary leak, disseminated intravascular coagulation, and disruption of the blood-brain barrier.

Algorithm

The researchers developed a predictive classification tree algorithm to identify patients who have an increased risk of severe neurotoxicity.

The algorithm suggests patients who meet the following criteria in the first 36 hours after CAR T-cell infusion have a high risk of grade 4-5 neurotoxicity:

• Fever of 38.9°C or greater
• Serum levels of IL6 at 16 pg/mL or higher
• Serum levels of MCP1 at 1343.5 pg/mL or higher.

This algorithm predicted severe neurotoxicity with 100% sensitivity and 94% specificity. Eight patients were misclassified, 1 of whom did not subsequently develop grade 2-3 neurotoxicity and/or grade 2 or higher CRS.

Funding

This research was funded by Juno Therapeutics Inc. (the company developing JCAR014), the National Cancer Institute, Life Science Discovery Fund, the Bezos family, the University of British Columbia Clinical Investigator Program, and via institutional funds from Bloodworks Northwest.

Dr Turtle receives research funding from Juno Therapeutics, holds patents licensed by Juno, and has pending patent applications that could be licensed by nonprofit institutions and for-profit companies, including Juno.

The Fred Hutchinson Cancer Research Center has a financial interest in Juno and receives licensing and other payments from the company.

Fred Hutch News Service

Researchers say they have identified potential biomarkers that may be used to help identify patients at an increased risk of neurotoxicity after chimeric antigen receptor (CAR) T-cell therapy.

The team also created an algorithm intended to identify patients whose symptoms were most likely to be life-threatening.

The researchers discovered the biomarkers and developed the algorithm based on data from a trial of JCAR014, an anti-CD19 CAR T-cell therapy, in patients with B-cell malignancies.

Cameron J. Turtle, MBBS, PhD, of Fred Hutchinson Cancer Research Center in Seattle, Washington, and his colleagues described this research in Cancer Discovery.

“It’s essential that we understand the potential side effects of CAR T therapies” Dr Turtle said. “While use of these cell therapies is likely to dramatically increase because they’ve been so effective in patients with resistant or refractory B-cell malignancies, there is still much to learn.”

Dr Turtle and his colleagues sought to provide a detailed clinical, radiological, and pathological characterization of neurotoxicity arising from anti-CD19 CAR T-cell therapy.

So the team analyzed data from a phase 1/2 trial of 133 adults with relapsed and/or refractory CD19+ B-cell acute lymphoblastic leukemia, non-Hodgkin lymphoma, or chronic lymphocytic leukemia.

The patients received lymphodepleting chemotherapy followed by an infusion of JCAR014.

Neurotoxicity

Within 28 days of treatment, 53 patients (40%) developed grade 1 or higher neurologic adverse events (AEs), 28 patients (21%) had grade 3 or higher neurotoxicity, and 4 patients (3%) developed fatal neurotoxicity.

Of the 53 patients with any neurologic AE, 48 (91%) also had cytokine release syndrome (CRS). All neurologic AEs in the 5 patients who did not have CRS were mild (grade 1) and transient.

Neurologic AEs included delirium with preserved alertness (66%), headache (55%), language disturbance (34%), decreased level of consciousness (25%), seizures (8%), and macroscopic intracranial hemorrhage (2%).

For most patients, neurotoxicity resolved by day 28 after CAR T-cell infusion. The exceptions were 1 patient in whom a grade 1 neurologic AE resolved 2 months after CAR T-cell infusion and the 4 patients who died of neurotoxicity.

The 4 neurotoxicity-related deaths were due to:

• Acute cerebral edema (n=2)
• Multifocal brainstem hemorrhage and edema associated with disseminated intravascular coagulation (n=1)
• Cortical laminar necrosis with a persistent minimally conscious state until death (n=1).

Potential biomarkers

In a univariate analysis, neurotoxicity was significantly more frequent in patients who:

• Had CRS (P<0.0001)
• Received a high CAR T-cell dose (P<0.0001)
• Had pre-existing neurologic comorbidities at baseline (P=0.0059).

In a multivariable analysis (which did not include CRS as a variable), patients had an increased risk of neurotoxicity if they:

• Had pre-existing neurologic comorbidities (P=0.0023)
• Received cyclophosphamide and fludarabine lymphodepletion (P=0.0259)
• Received a higher CAR T-cell dose (P=0.0009)
• Had a higher burden of malignant CD19+ B cells in the bone marrow (P=0.0165).

The researchers noted that patients who developed grade 3 or higher neurotoxicity had more severe CRS (P<0.0001).

“It appears that cytokine release syndrome is probably necessary for most cases of severe neurotoxicity, but, in terms of what triggers a person with cytokine release syndrome to get neurotoxicity, that’s something we need to investigate further,” said study author Kevin Hay, MD, of Fred Hutchinson Cancer Research Center.

Dr Hay and his colleagues also found that patients with severe neurotoxicity exhibited evidence of endothelial activation, which could contribute to manifestations such as capillary leak, disseminated intravascular coagulation, and disruption of the blood-brain barrier.

Algorithm

The researchers developed a predictive classification tree algorithm to identify patients who have an increased risk of severe neurotoxicity.

The algorithm suggests patients who meet the following criteria in the first 36 hours after CAR T-cell infusion have a high risk of grade 4-5 neurotoxicity:

• Fever of 38.9°C or greater
• Serum levels of IL6 at 16 pg/mL or higher
• Serum levels of MCP1 at 1343.5 pg/mL or higher.

This algorithm predicted severe neurotoxicity with 100% sensitivity and 94% specificity. Eight patients were misclassified, 1 of whom did not subsequently develop grade 2-3 neurotoxicity and/or grade 2 or higher CRS.

Funding

This research was funded by Juno Therapeutics Inc. (the company developing JCAR014), the National Cancer Institute, Life Science Discovery Fund, the Bezos family, the University of British Columbia Clinical Investigator Program, and via institutional funds from Bloodworks Northwest.

Dr Turtle receives research funding from Juno Therapeutics, holds patents licensed by Juno, and has pending patent applications that could be licensed by nonprofit institutions and for-profit companies, including Juno.

The Fred Hutchinson Cancer Research Center has a financial interest in Juno and receives licensing and other payments from the company.