Preemptive mitigation of CD19 CAR T-cell cytokine release syndrome without attenuation of antileukemic efficacy

Abstract

Immunotherapy with the adoptive transfer of T cells redirected with CD19-specific chimeric antigen receptors (CARs) for B-lineage acute lymphoblastic leukemia (ALL) can salvage >80% of patients having relapsed/refractory disease. The therapeutic index of this emerging modality is attenuated by the occurrence of immunologic toxicity syndromes that occur upon CAR T-cell engraftment. Here, we report on the low incidence of severe cytokine release syndrome (CRS) in a subject treated with a CAR T-cell product composed of a defined ratio CD4:CD8 T-cell composition with a 4-1BB:zeta CAR targeting CD19 who also recieved early intervention treatment. We report that early intervention with tocilizumab and/or corticosteroids may reduce the frequency at which subjects transition from mild CRS to severe CRS. Although early intervention doubled the numbers of subjects dosed with tocilizumab and/or corticosteroids, there was no apparent detrimental effect on minimal residual disease-negative complete remission rates or subsequent persistence of functional CAR T cells compared with subjects who did not receive intervention. Moreover, early intervention therapy did not increase the proportion of subjects who experience neurotoxicity or place subjects at risk for infectious sequelae. These data support the contention that early intervention with tocilizumab and/or corticosteroids in subjects with early signs of CRS is without negative impact on the antitumor potency of CD19 CAR T cells. This intervention serves to enhance the therapeutic index in relapsed/refractory patients and provides the rationale to apply CAR T-cell therapy more broadly in ALL therapy. This trial was registered at www.clinicaltrials.gov as #NCT020284.

Graphical abstract

Figure 1.

Treatment of CRS in the EI cohort with tocilizumab and dexamethasone. Persistent symptoms of mild CRS with a focus on persistent fever, early hypotension, and mild hypoxia are treated with tocilizumab (TOCI). Dexamethasone (DEX) is administered for more severe symptoms and continued until resolution of the symptoms. *For subjects who meet criteria for dexamethasone but have not yet received tocilizumab or received it >48 hours prior, tocilizumab is given concurrently with the dexamethasone.

Figure 2.

The onset of CRS following CAR T-cell infusion is variable and related to dose level with primary symptom of fever, which may be affected by increased use of tocilizumab in the EI cohort. (A) Swimmer plot of time to onset of CRS and its severity and duration (gray, no CRS; yellow, mild CRS; red, sCRS), with administration of tocilizumab (arrowheads) and dexamethasone (asterisks), as well as annotation for the assigned intervention cohort (DLT and EI; y-axis). (B) The onset of CRS by dose level of CAR T cells, as well as the severity of CRS by dose level (gray, mild CRS; red, sCRS). (C) Maximum fever following CAR T-cell infusion through day +28 grouped based on the presence or absence of sCRS. (D) Maximum fever based on CRS severity in the DLT cohort (D) and the EI cohort (E). *P < .05, **P  ≤  .01. NS, not significant.

Figure 3.

The relationship between laboratory values and severity of CRS. Absolute maximum values of CRP (A), ferritin (B) and LDH (C) between day 0 and day 28 in patients with sCRS or without CRS. over time from day 0 through day 28. Absolute neutrophil count (D), absolute lymphocyte count (E), platelet count (F), and hemoglobin levels (G) in patients with sCRS (red) and without CRS (blue) from day 0 to day 28. *P < .05, NS, not significant.

Figure 4.

Serum cytokines are impacted by administration of tocilizumab, regardless of the occurrence of sCRS. (A) GMCSF; (B) TNFα; (C) INFγ; (D) IL-2; (E) IL-5; (F) IL-6; (G) IL-10; (H) IL-13; (I) granzyme B; (J) granzyme A; (K) MIP-1b; (L) sCD137; (M) sFas. Serum cytokines in the peripheral blood are noted by the maximum value following CAR T-cell infusion through day 28, grouped by patients who did and did not develop sCRS. *P ≤ .05, **P  ≤  .01, ***P  ≤  .001, ****P ≤ .0001.

Figure 5.

LFS and OS are unaffected by CRS treatment strategy or administered CRS-directed treatment. All subjects were analyzed for LFS and OS following CAR T-cell infusions. The DLT cohort (red) and the EI cohort (blue) had similar LFS Kaplan-Meier curves (A) and OS Kaplan-Meier curves (B). Separately, all subjects were categorized per the specific intervention received. Intervention groups are steroid with or without tocilizumab (green), tocilizumab (blue), and none (red), with no difference between LFS Kaplan-Meier curves (C) and OS Kaplan-Meier curves (D).

Figure 6.

EI did not impact engraftment, expansion, or persistence of CAR T cells. All subjects were analyzed for engraftment and expansion of CAR T cells over time in the peripheral blood using flow cytometry. For analysis, all subjects were compared based on the occurrence of sCRS, on the intervention cohort assignment (DLT vs EI), and on the treatment received, regardless of the intervention cohort assignment. (A) The peak number of CAR T cells in the blood was higher in subjects who developed sCRS compared with those who did not. (B) The peak number of CAR T cells was not different between the intervention cohorts. (C) The peak number of CAR T cells was different among the 3 groups based on the immunomodulatory treatment received. (D) The area under the curve of CAR T-cell engraftment was higher in the group of subjects that experienced sCRS (red) vs no sCRS (blue). (E) The area under the curve of CAR T-cell engraftment was similar in the 2 intervention cohorts (red, DLT cohort; blue, EI cohort). (F) The area under the curve of CAR T-cell engraftment was enhanced in the groups that recieved immunomodulatory interventions (red, no intervention; blue, tocilizumab; green, steroids with or without tocilizumab). (G) The probability of having ongoing functional persistence of CAR T cells, as measured by BCA, is demonstrated by Kaplan-Meier curves. There was no difference between patients who developed sCRS (red) and those who did not (blue). There was no difference between the DLT (red) and EI cohorts (blue) (H) or among the immunomodulatory interventions received (red, no intervention; blue, tocilizumab; and green, steroids with or without tocilizumab) (I). NS, not significant; toci, tocilizumab.