Introduction: One of the main complications of adoptive T cell therapy (ACT) is the en-masse activation of tumor-reactive T cells inducing a large release of cytokines followed by activation of other immune cells leading to adverse events. These are classified as a cytokine release syndrome (CRS) or neurotoxicity described as a CAR T Related Encephalopathy Syndrome (CRES). Several biomarkers have been associated with CRS and/or neurotoxicity such as LDH, ferritin and CRP. Cytokines have also been associated with CRS and and/or CRES, but present approaches rely on retrospective study of collected biomarkers. Here, we report the results of cytokine analysis using a point of care (POC) device to predict immune-related toxicities in patients with relapsed/refractory (R/R)DLBCL treated with axicabtagene ciloleucel (axi-cel).

Methods: Patients with R/R DLBCL treated with commercial axi-cel were included in this study. Baseline serum samples were collected prior to lymphodepleting chemotherapy and then daily during hospitalization. To select which cytokines to monitor, we retrospectively analyzed 38 serum cytokines in a cohort of 53 patients with R/R B cell acute lymphoblastic leukemia (B-ALL) who were treated with 19-28z CAR T cells. The patients were divided into those requiring treatment with tocilizumab and/or steroids versus those who did not require treatment. We observed several cytokines, including IL-2, IL-6, IL-15 and IFNg, which were significantly elevated in patients with CRS and/or CRES requiring treatment (Figure 1a). Based on this analysis and results of published studies, eight serum proteins were selected in our study including IL-1b, IL-2, IL-6, IL-15, IFNg, TNFa, and angiopoietin-1 &2. We monitored these proteins using a POC device that allows for rapid daily monitoring with a turnaround time of two hours. We established that the results from the POC device strongly correlate with a current gold standard device(Luminex), which has a typical two day turn around time. CRS and CRES were prospectively graded using revised Lee criteria (Lee et al Blood 2014) and the CARTOX group (Neelapu et al. NRCO 2017) respectively by an experienced clinical team and confirmed by chart review retrospectively.

Results: A total of 20 patients with R/R DLBCL treated with commercial axi-cel were identified. Median age 64 years ( range 43-73) with 80% male.In our cohort, grades 1-3 CRS were observed in 45%, 40% and 5% respectively. There were no observed grade 0 or grade 4 CRS. There were two patients (10%) who died in the setting of severe toxicity. Patients with grade 5 CRS had higher levels of IL-6 and angiopoietin 2/angiopoietin 1 ratio at day one, which correlated with severity of toxicity r=0.52 (p= 0.039) , and r=0.53 (p=0.033) respectively (Fig. 1b). Furthermore, patients with high grades CRS had elevated levels of IL-15 at day seven (r=0.83, p=0.006). The majority of patients (55%) had grade 1-2 CRES.There were no significant correlations between serum cytokine levels and CRES or between those who required tocilizumab/steroids vs. those who did not, likely due to the small sample size. In select cases, daily monitoring of cytokines using the POC device provided clinical insight that wasn't evident from standard biomarkers. For example, one patient who developed delayed CRS had high serum levels of IL-6 but did not have elevated levels of CRP(Fig.1c).

Discussion: In this analysis of 20 patients, we observed a correlation between severe CRS and elevated serum cytokine levels of IL-6 and angiopoietin 2/angiopoietin 1 ratio at day one suggesting that these biomarkers may be utilized to predict severe toxicity in patients treated with ACT. While this study is limited by small sample size, our observations correlate with previously published biomarkers data in patients enrolled in clinical trials. To our knowledge this is the first reported cytokine data using commercial axi-cel. Monitoring of cytokines using a POC device is feasible and will be useful clinically. High risk patients may be identified early and help guide intervention in real time, for example day one elevated IL-6 levels might inform earlier use of tocilizumab. We continue to enroll patients to validate cytokines as predictive biomarkers with the goal of informing the development of preventative strategies to mitigate CAR T cell therapy immune related adverse events.


Locke:Cellular BioMedicine Group Inc.: Consultancy; Kite Pharma: Other: Scientific Advisor; Novartis Pharmaceuticals: Other: Scientific Advisor. Brentjens:Juno Therapeutics, a Celgene Company: Consultancy, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties, Research Funding. Park:Amgen: Consultancy, Membership on an entity's Board of Directors or advisory committees; AstraZeneca: Consultancy; Kite Pharma: Consultancy; Juno Therapeutics: Consultancy, Research Funding; Adaptive Biotechnologies: Consultancy; Novartis: Consultancy; Pfizer: Consultancy; Shire: Consultancy. Davila:Celyad: Consultancy, Membership on an entity's Board of Directors or advisory committees.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution