In this issue of Blood, Karschnia et al spotlight the “Achilles heel” of chimeric antigen receptor (CAR) T-cell therapy and call for prospective clinical trials to evaluate strategies to manage and potentially prevent CAR T cell–induced neurotoxicity (NT).1 

Pathophysiologic mechanism of neurotoxicity and CRS. GM-CSF is produced by CAR T cells upon contact with tumor and serves as a communication conduit between the specific immune response of the CAR T cells and the off-target inflammatory cascade produced by myeloid lineage cells. GM-CSF acts directly on myeloid cells to expand, activate, and promote the production of other chemokines, including MCP-1/CCL2, IP-10/CXCL10, and cytokines IL-6 and IL-1. Once initiated, the inflammatory cascade can become self-perpetuating as the production of chemokines results in further expansion and trafficking of myeloid cells to the tumor bed. The positive feedback loop can result in abnormally high levels of inflammatory cytokines, endothelial activation, vascular permeability, and ultimately, NT and CRS. GM-CSF also acts directly on myeloid lineage cells to promote the expansion and trafficking of myeloid derived suppressor cells (MDSC) and tumor-associated macrophages (TAM), which have been demonstrated to inhibit T-cell proliferation and effector functions.

Pathophysiologic mechanism of neurotoxicity and CRS. GM-CSF is produced by CAR T cells upon contact with tumor and serves as a communication conduit between the specific immune response of the CAR T cells and the off-target inflammatory cascade produced by myeloid lineage cells. GM-CSF acts directly on myeloid cells to expand, activate, and promote the production of other chemokines, including MCP-1/CCL2, IP-10/CXCL10, and cytokines IL-6 and IL-1. Once initiated, the inflammatory cascade can become self-perpetuating as the production of chemokines results in further expansion and trafficking of myeloid cells to the tumor bed. The positive feedback loop can result in abnormally high levels of inflammatory cytokines, endothelial activation, vascular permeability, and ultimately, NT and CRS. GM-CSF also acts directly on myeloid lineage cells to promote the expansion and trafficking of myeloid derived suppressor cells (MDSC) and tumor-associated macrophages (TAM), which have been demonstrated to inhibit T-cell proliferation and effector functions.

Close modal

Although the emergence of CAR T-cell therapy has dramatically improved response rates for patients with relapsed or refractory B-cell hematologic malignancies, its utility is hampered by the potential for significant side effects, including severe NT, which as reported by Karschnia et al is a negative prognostic factor for overall survival. Corticosteroids are currently the recommended treatment of NT, but prolonged exposure of >10 days in patients with severe NT may negatively influence overall survival. The potential impact of NT on overall survival is a significant finding as >50% of CAR T cell–treated patients with NT in this study had developed severe NT. In addition, NT is associated with cytokine release syndrome (CRS), and the anti–interleukin-6 (IL-6) receptor antagonist, tocilizumab (Actemra), currently the only Food and Drug Administration (FDA)-approved therapy for the treatment of severe CRS, has been shown to increase both the overall rate of NT and the rate of severe NT when used prophylactically.2  Moreover, the majority of patients treated with CAR T cell are treated as in-patients, and admission to the intensive care unit (ICU) for the management of these toxicities is often required, creating an added health economic burden and less favorable reimbursement for hospitals and institutions, which inevitably results in restricted access. Analyses of health resource utilization point to length of hospitalization and length of stay in the ICU as the primary drivers of non-drug–related costs for CAR T cell–treated patients, which are projected to be at least twice as high for those who develop these severe toxicities.3,4  Strategies to improve the safety profile of CAR T-cell therapy without negatively impacting efficacy are needed to improve its benefit-to-risk profile and cost-effectiveness and to enable CAR T-cell therapy to move beyond use solely in relapsed/refractory patients to earlier lines of therapy.

There are no FDA-approved therapies available for the prevention, nor for the treatment, of NT. Much has been learned regarding the possible mechanisms and pathophysiology of CAR T cell–induced NT, including the role of myeloid cells, endothelial cells, and proinflammatory cytokines. In addition to ferritin, a biomarker that has been shown by Karschnia et al and others to correlate with severe NT, an analysis by Rossi et al evaluating axicabtagene ciloleucel (axi-cel, Yescarta), the first CAR T cell therapy approved for the treatment of relapsed/refractory diffuse large B-cell lymphoma, showed that levels of IL-15 and granulocyte-macrophage colony-stimulating factor (GM-CSF) are elevated 1 day following CAR T-cell administration, and the early elevation of these cytokines are correlated with severe NT.5  No other proinflammatory cytokines were directly or indirectly associated with severe NT. It is proposed that upon contact with the tumor, CAR T cells produce GM-CSF,6  which serves as a communication conduit between the specific immune response of the CAR T cells and the off-target inflammatory cascade produced by myeloid lineage cells, causing myeloid cells to expand and promote the production of other proinflammatory chemokines and cytokines, including monocyte chemoattractant protein-1 (MCP-1), IL-1, and IL-6, among others (see figure). Fever and elevated levels of MCP-1 36 hours after CAR T-cell administration have been demonstrated to be the best predictors of severe NT and CRS with high levels of specificity and sensitivity.7  Moreover, IL-6 is only released by the antigen-presenting cells, or tumor cells, in a contact-independent manner, which helps explain why the prophylactic administration of tocilizumab is not effective in reducing the overall incidence of CRS or NT, as this cytokine is downstream in the inflammatory cascade.2,6  Once initiated, this inflammatory cascade can become a self-perpetuating “storm,” resulting in further expansion and trafficking of myeloid cells to the tumor bed, abnormally high levels of inflammatory cytokines, endothelial activation, vascular permeability, and ultimately, CRS and NT.

Blood-brain barrier (BBB) disruption and infiltration of myeloid cells and proinflammatory cytokines into the central nervous system (CNS) are other important factors in the pathogenesis of NT. In the study by Karschnia et al, low platelet counts, a biomarker for BBB disruption, prior to CAR T-cell infusion were associated with severe NT. The integrity of the BBB can be noninvasively monitored by magnetic resonance imaging (MRI). Conventional contrast agents containing gadolinium are used in association with MRI to detect and quantify BBB leakage. Preclinical in vivo studies have shown diffuse neuroinflammation and BBB impairment following CAR T-cell therapy, enabling a massive influx of proinflammatory cytokines into the CNS, which is thought to propagate neuroinflammation.8  This is consistent with data reported in CAR T cell clinical trials, where disruption of the BBB and a significant increase in levels of proinflammatory cytokines and CD14+ myeloid cells in the cerebrospinal fluid were seen in patients who developed severe NT, suggesting the potential for local CNS-specific production.2,9 

Considering the high rates of NT, including severe cases, after CAR T-cell therapy, strategies to manage or prevent the onset of NT need to be evaluated in prospective clinical trials as proposed by Karschnia et al. One such strategy that has shown significant promise is GM-CSF neutralization, which for the first time has demonstrated that the toxicities associated with CAR T-cell therapy can be effectively prevented in vivo.8  Preclinical in vivo studies have shown that neutralizing GM-CSF prevents CAR T cell–induced CRS and significantly reduces NT.8,10  Quantification of MRI via gadolinium-enhanced T1 hyperintensity showed a 75% decrease in neuroinflammation and BBB impairment. Enhanced antitumor activity, improved overall survival, and improved durability of response with a reduced rate of relapse were also observed with GM-CSF neutralization in this xenograft model.8  Prospective clinical trials evaluating GM-CSF neutralization in combination with CAR T-cell therapy are expected to be initiated this year. Other approaches to improve the safety of CAR T-cell therapy include the use of debulking chemotherapy, and high-dose corticosteroids; however, the extent of improvement in rates of NT and CRS and whether this comes at the expense of efficacy remain to be determined.

Conflict-of-interest disclosure: O.A. is employed by Humanigen.

1.
Karschnia
P
,
Jordan
JT
,
Forst
DA
, et al
.
Clinical presentation, management, and biomarkers of neurotoxicity after adoptive immunotherapy with CAR T cells
.
Blood
.
2019
;
133
(
20
):
2212
-
2221
.
2.
Locke
FL
,
Neelapu
SS
,
Bartlett
NL
, et al
.
Preliminary results of prophylactic tocilizumab after axicabtagene ciloleucel (axi-cel; KTE-C19) treatment for patients with refractory, aggressive non-Hodgkin lymphoma (NHL) [abstract]
.
Blood
.
2017
;
130
(
suppl 1
).
Abstract 1547
.
3.
Hernandez
I
,
Prasad
V
,
Gellad
WF
.
Total costs of chimeric antigen receptor T-cell immunotherapy
.
JAMA Oncol
.
2018
;
4
(
7
):
994
-
996
.
4.
Siddiqi
T
,
Garcia
J
,
Dehner
C
, et al
.
Estimation of the resource utilization and costs of cytokine release syndrome observed in the transcend-NHL clinical trial: a micro-costing study [abstract]
.
Blood
.
2018
;
132
(
suppl 1
).
Abstract 319
.
5.
Rossi
JM
,
Sherman
JM
,
Xue
A
, et al
. Low-dose conditioning chemotherapy and anti-CD19 CAR T cells may elicit distinct immune programs associated with clinical responses. EMA Workshop. Nov 15-16, 2016
6.
Barrett
DM
,
Singh
N
,
Hofmann
TJ
, et al
.
Interleukin 6 is not made by chimeric antigen receptor T cells and does not impair their function [abstract]
.
Blood
.
2016
;
128
(
22
).
Abstract 654
.
7.
Gust
J
,
Hay
KA
,
Hanafi
LA
, et al
.
Endothelial activation and blood-brain barrier disruption in neurotoxicity after adoptive immunotherapy with CD19 CAR-T cells
.
Cancer Discov
.
2017
;
7
(
12
):
1404
-
1419
.
8.
Sterner
RM
,
Sakemura
R
,
Cox
MJ
, et al
.
GM-CSF inhibition reduces cytokine release syndrome and neuroinflammation but enhances CAR-T cell function in xenografts
.
Blood
.
2019
;
133
(
7
):
697
-
709
.
9.
Santomasso
BD
,
Park
JH
,
Salloum
D
, et al
.
Clinical and biological correlates of neurotoxicity associated with CAR T-cell therapy in patients with B-cell acute lymphoblastic leukemia
.
Cancer Discov
.
2018
;
8
(
8
):
958
-
971
.
10.
Sachdeva
M
,
Duchateau
P
,
Depil
S
, et al
.
Granulocyte-macrophage colony stimulating factor inactivation in CAR T-Cells prevents monocyte-dependent release of key cytokine release syndrome mediators
.
J Biol Chem
.
2019
.
294
:
5430
-
5437
.