Adaptable Leukemia Cells Resisting CAR T-cell Attack via B-cell Activation

The targeting of blood cancers using chimeric antigen receptor (CAR)-engineered T cells has brought deep, durable remissions and even cures to some patients who have relapsed, refractory B-cell malignancies. CARs are synthetic receptors made up of an antigen-binding portion of an antibody (e.g., anti-CD19) and critical signaling domains from T-cell proteins such as CD28, 4–1BB, and CD3. Current FDA-approved second-generation CAR T cells express either a CD28 or 4–1BB cosignaling receptor.

Despite impressive clinical responses, escape from CAR T-cell therapy occurs frequently and understanding the mechanisms underlying this is an active area of research. Antigen loss or downmodulation is one of the most common causes of tumor escape following CAR T-cell therapy (1–3), yet resistance to CAR T cell–mediated lysis has also been reported (4). Antigen loss or downregulation can occur through numerous processes, including antigen masking, internalization, and trogocytosis, but the essence of how tumor cells survive CAR T-cell therapy has remained elusive.

Using Nalm-6 as a model of acute lymphoblastic leukemia (ALL) and third generation CD19-targeting CAR T cells with both CD28 and 4–1BB co-stimulatory domains, Im and colleagues have unveiled features of tumor cells resistant to CAR T-cell killing, finding that early interactions between CAR T cells and tumor cells can enable escape of some tumor cells (5). At the point of initial contact of CAR T cells with tumor cells, the authors show, using live cell imaging, that there are three patterns of interactions: Effective cytotoxic killing, “scanning interactions”, and “escape interactions”. The interactions are characterized by differential synaptic dwelling times. Visual tracking of the CD19 antigen on surviving leukemic cells demonstrates that formation of the immunologic synapse triggers CD19 clustering, internalization, and degradation. Thus, surface-antigen expression on tumor cells decreases upon exposure to CAR T cells. More importantly, CD19^low NALM6 cells become resistant to fresh CAR T cell–mediated killing.

To go back to the original question: Which tumor cells got spared by CAR T cells? Through single-cell RNA sequencing, the authors demonstrate that leukemic-cell escape from CAR T-cell killing is partially due to transcriptionally-driven decreases in CD19 expression and engagement of B-cell activation signatures. These signatures are characterized by CD69, CCR7, and NFκB-, and CD40-related pathways, and lead to transcriptional and epigenetic changes associated with a germinal center reaction.

Of potential translational interest, pharmacologic inhibition of the tyrosine kinase BTK, to inhibit B-cell activation, is shown to boost killing of NALM6 cells by CAR T cells. These data suggest that the inhibition of leukemic-cell activation via the B-cell receptor blunts prosurvival pathways in this cell line. More generally, the result suggests that activated B cells may have better odds of escaping CAR T-cell killing. As with all good studies, the findings reported by Im and colleagues raise some open-ended questions for the CAR T-cell field: Will this be the same in ALL patient samples? Can this conclusion be generalized to other B-cell malignancies? How can we apply this finding to improve therapy?

As we learn more about leukemic-cell plasticity under conditions of CAR T-cell therapy, through studies such as that by Im and colleagues, we will be able to predict treatment outcomes more accurately and identify new potential treatment maneuvers to prevent leukemic cells evading CAR T-cell surveillance.

J.J. Melenhorst reports personal fees from IASO Biotherapeutics, Poseida Therapeutics; and personal fees from Kite Pharma outside the submitted work; in addition, J.J. Melenhorst has a patent for methods for improving the efficacy and expansion of immune cells issued to Novartis, a patent for Biomarkers predictive of therapeutic responsiveness to chimeric antigens issued to Novartis, a patent for methods of making chimeric antigen receptor - expressing cells issued to Novartis, and a patent for methods for improving the efficacy and expansion of chimeric antigen receptor issued to Novartis. No disclosures were reported by the other author.