CAR emissions: cytokines tell the story

In this issue of Blood, Hay et al describe the cytokine profile associated with cytokine release syndrome (CRS) in a large cohort of adults treated with CD19-directed chimeric antigen receptor (CAR)-modified T cells.¹ 

In recent years, clinical trials of CD19 CAR T cells have had a transformative impact on the treatment of multiply relapsed and refractory B-cell malignancies, once thought to be incurable.^2-4  These trials paved the way for the first-ever US Food and Drug Administration approval of a cellular gene therapy, in August 2017 with the approval of Kymriah (tisagenlecleucel, formerly CTL019), a CD19-targeted CAR T-cell product developed by Novartis in collaboration with the University of Pennsylvania (Penn) and Children’s Hospital of Philadelphia (CHOP). With this landmark approval, this once highly specialized technology, formerly available at only a few institutions, will be more widely available across the United States. Yet, the care required to safely administer this therapy remains highly specialized.

CRS is an expected, potentially serious, and sometimes life-threatening toxicity reported by all groups who have demonstrated exponential proliferative capacity and high response rates of CD19 CAR T cells.^1-6  Here, Hay and colleagues describe the cytokine profile of a large cohort (133) of adults with acute lymphoblastic leukemia (ALL), chronic lymphocytic leukemia (CLL), and non-Hodgkin lymphoma (NHL) treated with the CD19 CAR T-cell product developed at the Fred Hutchinson Cancer Research Center in Seattle. Within the first 36 hours after infusion, higher levels of interferon-γ (IFN-γ), interleukin-6 (IL-6), IL-8, IL-10, monocyte chemoattractant protein-1 (MCP-1), tumor necrosis factor receptor p55, and macrophage inflammatory protein-1β were associated with development of grade 4/5 CRS (see figure). These associations are consistent with those reported by Teachey et al in patients with ALL treated with CTL019 at CHOP and Penn, and point to the key involvement of macrophages.⁷  Both reports identified prediction models of CRS incorporating early cytokine measurements. In the current report, Hay et al demonstrated that a fever ≥38.9°C combined with a serum MCP-1 concentration ≥1343.5 pg/mL within 36 hours after infusion predicted grade 4/5 CRS in their cohort with a sensitivity of 1.00 and a specificity of 0.95. A novel finding of this study is the identification of elevations in von Willebrand factor, angiopoietin-2 (Ang-2), and the Ang-2:Ang-1 ratio in patients with severe CRS. Rises in these markers of endothelial activation suggest this may be an important mechanism in the development of severe CRS.

This work highlights key similarities in CRS associated with distinct CAR technologies: other groups have also observed the association of severe CRS with high bone marrow tumor burden, liver and kidney dysfunction, coagulopathy, macrophage activation syndrome, and early fever.^2-4,7  Perhaps more importantly, this work highlights key differences in CRS grading. In this study, Hay et al use the National Cancer Institute (NCI) CRS grading scale⁸ ; in their study, Teachey et al used the Penn CRS grading scale.⁹  When evaluating toxicities of CAR T-cell trials, cytokines associated with higher-grade CRS, and CRS prediction models, an understanding of the CRS grading scale used and the differences between grading scales is vital. A key difference involves vasopressor use: on the NCI CRS grading scale, a patient requiring high-dose vasopressors could be classified as grade 3, whereas the same patient would be classified as grade 4 on the Penn scale. This discrepancy makes it challenging to compare toxicities and cytokine profiles across trials.

As CD19 CAR T-cell therapies become more widely available, recognition of risk factors, and ideally predictors, of severe CRS will be critical for clinicians to safely administer this therapy. The power of this study is the size and breadth of the cohort examined. This work demonstrated similar cytokine profiles after CD19 CAR T-cell infusion in patients with ALL, CLL, or NHL, a previously unclear parallel. Several unanswered questions remain. (1) Is the cytokine pattern and time course comparable across CAR designs? (2) Are CRS prediction models generalizable? (3) Will early intervention mitigate serious complications of CRS? Collaborative studies of cytokine profiles across trials and CAR designs as well as in larger populations will be vital to a complete understanding of CRS. Ongoing trials incorporating CRS prediction models or early intervention with cytokine blockade will add key insights into toxicity management.