B-cell activating factor plays a critical role in CAR-T cell-associated cytokine release syndrome

Abstract

Cytokine release syndrome (CRS) is the most common and potentially life-threatening toxicity associated with CAR-T cell therapy and is related to a heightened immune effector state. In this work, we identified a novel role of the pro-tumorigenic cytokine B-cell activating factor (BAFF) in its pathophysiology. First, we observed that patients who experienced CAR-T cell-related CRS have elevated serum BAFF levels that coincide with elevated IL-6. Mechanistically, we show that IFN-γ, produced by activated CAR-T cells, stimulates monocytes to release BAFF, which induces the expression of CRS-related cytokines from monocytes. Monocytes derived from CRS patients express BCMA, which is further induced by IFN-γ stimulation. Neutralization of BAFF with belimumab significantly reduces production of various CRS and ICANS-related cytokines without impairing CAR-T cell activation or killing. Overall, we demonstrate that BAFF plays a critical role in CAR-T-cell-related CRS, and its neutralization may be a novel strategy for treating both CRS and ICANS.

Figure 1:. Serum BAFF is elevated in CAR-T patients who experience CRS.

Serum from CRS+ patients at various timepoints after CAR-T-cell infusion (D=day post CAR-T-cell infusion) was subjected to Multiplex or Luminex cytokine analysis. ‘PT’ is de-identified patient number.

Figure 2:. Serum BAFF does not increase in patients who did not experience CAR-T related CRS.

Serum from patients who received CAR-T cell therapy and were not diagnosed with CRS at various timepoints after CAR-T-cell infusion (D=day post CAR-T-cell infusion) was subjected to Multiplex or Luminex cytokine analysis.

Figure 3.. Contributions of cell types to cytokine release.

Jeko-1 cells, CD19 CAR-T cells, and THP-1 cells were cultured alone or in every combination at a 1:1:1 ratio for 24h, when the supernatant was harvested for cytokine analysis by Multiplex flow cytometry. This data represents 2 independent experiments.

Figure 4.. IFN-γ stimulates BAFF secretion and upregulation of BCMA on monocytes. BAFF stimulates expression of CRS cytokines in monocytes.

THP-1 were incubated with 50ng/ml rhlFNγ. After 48h supernatant was collected and subject to (A) multiplex analysis to detect secreted BAFF (p=0.0122) (data represents 3 independent experiments) and (B) the cells were stained with a panel of antibodies for BCMA(PE), TACl(Pcy7), and BAFF-R (APC) and analyzed via flow cytometry. CD14+ monocytes were isolated from (C) healthy donor (n=6) or (D) CAR-T-CRS patient-derived (n=2) PBMCs and were stimulated and analyzed as in (A). E THP-1 cells were stimulated with 250ng/ml of rhBAFF. After various timepoints, RNA was isolated, and expression of various cytokines was measured by RT-qPCR (this data represents 2 independent experiments).

Figure 5.. BAFF neutralization reduces cytokine release.

(A) Jeko-1 cells were pre-incubated with 100μg/ml belimumab for 24h before co-culture with THP-1 and CD19 CAR-T cells (“−”: no belimumab). After another 16–24 the supernatant of the co-culture was harvested and subject to cytokine analysis by Multiplex. An unpaired t-test was used to compare untreated and belimumab conditions for each analyte. This data represents three independent experiments. P-values: BAFF 0.0017, IL-6 0.0234, IFN-γ 0.0004, APRIL 0.0161, CXCL8 0.0029, CCL2 0.88454, GM-CSF 0.0004, IL-1β 0.037, IL-10 0.0013. (B) SCIO beige mice were inoculated intravenously with 3 million Raji-luciferase cells. 3 mice were treated with atacicept every 2 days starting at day 11 post inoculation. At day 15 6 mice received 10 million CD19 CAR-T cells. Serum was harvested at day 17 and subjected to murine multiplex analysis.

Figure 6.. BCMA neutralization reduces cytokine release.

THP-1 cells were pre-incubated with 10μg/ml of anti-BCMA neutralizing antibody for 24h. THPs were then co-cultured with Jeko-1 cells and CD19 CAR-T cells at a ratio of 5:1:5. Supernatant from the co-culture was harvested 8 or 24 hours and subject to cytokine analysis by Multiplex. This data represents 2 independent experiment. P-values: BAFF 0.0217, IL-6 0.0085, IFN-γ 0.0504, APRIL 0.0180, CXCL8 0.0002, CCL2 0.0128, GM-CSF 0.0388, IL-1β 0.0182, IL-10 0.0158.

Figure 7.. BAFF neutralization does not affect CAR-T cell function or activation.

(A) Jeko-1 cells were stained with efluoro670 dye and pre-incubated with 100μg/ml belimumab (“Bel”) for 24h before co-culture with THP and CD19 CAR-T cells for an additional 16–20h. Cells were stained with propidium iodide and % APC and Pl – positive cells were reported. This data represents data from 2 CAR-T cell donors and 2 independent experimental replicates of each. (B) Raji cells were pre-incubated with 100μg/ml belimumab before co-culture with CAR-T and THP cells. After 16–20 additional hours of co-culture, cells were stained with CD3 (PE) and CD69 (Percp) and the percentage of CD3+ cells that are CD69+ was reported. (C) Raji cells were pre-incubated with 100μg/ml belimumab before co-culture with CAR-T and THP cells. Cells were immediately stained with CD107a (Pcy7). After 1 hour monensin golgi was added and cells were incubated for another 4–5 hours and were then stained with CD3 (PE). Percent positive CD3+ cells that stained with CD107a were reported for degranulation. The experiments in panels (B) and (C) represent 3 donors of CAR-T cells and 2 independent experiments per donor.