AUTHORS
Nat Murren, BS, Isaac King2, Lauren Mahoney3, Jarron Roy4, Otto Kletzien, PhD5, Maddy Collins6, Sarah Geffe7, Luke Kalcheim8, Rebecca Richards, MD, PhD9
AFFILIATIONS
1 University of Wisconsin-Madison, Madison, Wisconsin, United States
2 University of Wisconsin-Madison, Madison, Wisconsin, United States
3 University of Wisconsin-Madison, Madison, Wisconsin, United States
4 University of Wisconsin-Madison, Madison, Wisconsin, United States
5 University of Wisconsin-Madison, Madison, Wisconsin, United States
6 University of Wisconsin-Madison, Madison, Wisconsin, United States
7 University of Wisconsin-Madison, Madison, Wisconsin, United States
8 University of Wisconsin-Madison, Madison, Wisconsin, United States
9 University of Wisconsin-Madison, Madison, Wisconsin, United States
ABSTRACT
Despite the success of chimeric antigen receptor (CAR) T cell therapy for B cell acute lymphoblastic leukemia (B-ALL), its translation to acute myeloid leukemia (AML) has been limited by poor efficacy and significant toxicity. Interferon-gamma (IFNγ) blockade with emapalumab has recently emerged as a promising strategy to mitigate CAR T cell-related toxicities in B cell malignancies, based on evidence that IFNγ is largely dispensable for optimal CAR T cell activity in B-ALL. Whether IFNγ signaling is similarly dispensable in the AML context remains unclear. In solid tumor models, disruption of IFNγ signaling has been associated with increased tumor resistance to CAR T cell killing, partially attributed to reduced expression levels of adhesion molecule ICAM-1. IFNγ is known to regulate expression of ICAM-1 and apoptotic death receptors like Fas, both of which are commonly expressed on AML. Thus, we hypothesized that unlike in B-ALL, an intact IFNγ axis is required for CAR T cells to effectively kill AML.
To investigate this, we engineered Nalm-6 B-ALL cells to express CD123 (N6-123) and AML cell lines to express CD19. These cell lines were co-cultured with CD19 or CD123 CAR T cells with and without IFNγ inhibition by pharmacologic blockade or IFNγ genetic knockout (IFNγKO) in serial re-challenge cytotoxicity assays. Tumor growth was quantified using Incucyte live-cell imaging. Neither pharmacologic blockade nor IFNγ knockout impacted CD123 CAR T cells cytotoxicity against B-ALL N6-123 cells. In contrast, disruption of IFNγ impaired both CD123 and CD19 CAR T cell killing of multiple AML cell lines. We validated these data in co-cultures with primary leukemia samples, using LDH release to quantify cytotoxicity. Again, IFNγ blockade significantly reduced CD123 CAR T cell cytotoxicity against primary AML but had no effect on CD19 CAR T cells against primary B-ALL. In vivo, NSG mice were engrafted with AML cell line OCI-AML3 then treated with wild-type (WT) or IFNγKO CD123 CAR T cells, and tumor burden was monitored with bioluminescence imaging. These data similarly demonstrated a trend toward IFNγ dependence, with mice treated with IFNγKO CAR T cells displaying reduced tumor control compared to WT CAR T cells.
Mechanistically, we hypothesized that IFNγ promotes AML susceptibility to CAR T cell killing by remodeling the tumor cell surface phenotype. Therefore, utilizing flow cytometry, we investigated the surface expression of CD123, ICAM-1, and Fas on AML and B-ALL cell lines both during co-culture and after exposure to AML-CAR T conditioned media. Improved cytotoxicity of CD123 CAR T cells against AML correlated with IFNγ-mediated CD123 and ICAM-1 upregulation during and after co-culture. Exposure to supernatant from CAR T cell co-cultures also induced CD123, ICAM-1, and Fas upregulation on AML cell lines. In both cases, the upregulation of these key molecules on AML was either significantly reduced (ICAM-1 and Fas) or completely diminished (CD123) by IFNγ blockade, whereas expression on N6-123 cells remained unaffected. To assess its contribution to CAR T cell cytotoxicity, ICAM-1 function was inhibited with a blocking antibody, alone or in combination with IFNγ blockade. ICAM-1 blockade significantly reduced CD123 CAR T cell killing of AML but not B-ALL, and combined blockade of IFNγ and ICAM-1 further exacerbated this effect. Finally, AML and B-ALL cell lines were incubated with exogenous IFNγ and co-cultured with CD123 CAR T cells for cytotoxicity assays and avidity measurements. The addition of exogenous IFNγ improved CAR T cell cytotoxicity and avidity against AML but not B-ALL, coinciding with increased expression of CD123, ICAM-1, and Fas expression only on AML.
Our data demonstrate that disruption of the IFNγ axis impedes anti-AML CAR T cell activity and prevents upregulation of target antigen CD123 and immune synapse and apoptotic mediators ICAM-1 and Fas on AML cells. These findings identify IFNγ as a critical mediator of CAR T cell efficacy in AML by promoting both target antigen expression and stable CAR T cell-tumor interactions. Our results suggest that therapeutic IFNγ blockade, including with emapalumab, may compromise CAR T cell responses in AML and should be approached with caution in this disease context. Altogether, our data highlight core differences between AML and B-ALL susceptibility to CAR T cells and highlight the importance of defining a tumor-specific molecular microenvironment.
DISCLOSURES
Nat Murren, BS: No relevant conflicts of interest to declare.