Development of a First-in-Class CAR-T Therapy Against Calreticulin-Mutant Neoplasms and Evaluation in the Relevant Human Tissue Environment

Introduction

Myeloproliferative neoplasms (MPNs) are characterized by a TGFb-rich, fibrotic tumor microenvironment (TME), posing a challenge for targeted immunotherapies. Current treatments are largely palliative and do not selectively target MPN clones. In 1/3 of patients, MPNs are driven by mutant calreticulin (mutCALR), that binds to the thrombopoietin receptor (TpoR) causing activation and prominent display of the cancer neoantigen on the surface of TpoR-expressing cells – stem/progenitors (HSPCs) and megakaryocytes. Bispecific and inhibitory antibodies targeting mutCALR recently entered clinical trials. Here, we present pre-clinical validation of a pioneering chimeric antigen receptor T-cell (CAR-T) therapy targeting mutCALR, including efficacy in relevant clinical settings.

Methods

Killing efficacy and target cell TpoR expression was assessed by flow cytometry. To evaluate CAR-T in the relevant TME, we used scRNAseq to analyze iPSC-derived human bone marrow (BM) organoids co-engrafted with CAR-T, patient-derived CD34+ HSPCs +/- immunoregulatory factors abundant in the myelofibrotic niche (TGFβ and Galectin-1,Gal1).

Results:

The novel CAR-T demonstrated robust and highly selective eradication of low- and high- mutCALR expressing human cell lines in vitro, and CAR-T administration led to a dramatic reduction in leukemic burden and improved survival in NSG mice xenografted with mutCALR+ TpoR+ AML cells. Excellent BM infiltration of CAR-T, and near-complete ablation of leukaemia confirmed remarkable efficacy of CAR-T against mutCALR+ malignancies in vivo.

CAR-T efficacy was tested on CD34+ HSPCs from MPN patients (n=12), including myelofibrosis, ET, and accelerated/blast phase MPN (AP/BP-MPN) with both ins5 and del52 variants (n=8) in 2D and 3D cultures. Excellent depletion of HSPCs (40->90%) and extreme specificity was observed, with minimal killing of JAK2V617F+ samples (n=4).

Killing of AP/BP-MPN HSPCs was notably lower than chronic phase samples (myelofibrosis/ET), and correlated with lower surface TpoR expression. We hypothesized that treatment with TpoR agonists might increase TpoR expression on AP/BP-MPN HSPCs, by promoting megakaryocytic differentiation. Serial doses of rhTPO and TpoR agonists were tested. Pre-treatment of cells with Eltrombopag, but not with Romiplostim or rhTPO, increased surface TpoR expression by >50% and improved CAR-T killing, confirming that pharmacologically enhancing TpoR expression might boost vulnerability to anti-mutCALR immunotherapies.

As expected, CAR-T killing efficacy was substantially diminished by prolonged exposure to JAK inhibitors at clinically-relevant concentrations, with a substantial drop in cytotoxicity and CAR-T proliferation. However, 24hr pre-incubation of CAR-T with peak plasma equivalent doses of JAK inhibitors prior to exposure to target cells led only to a smaller reduction in killing, ranging from <10-40%. Less potent JAK1 inhibitors demonstrating only a minor drop in efficacy, suggesting that selected JAKi agents/dosing regiments or short washout periods may be possible for synergistic co-therapy with CAR-T treatment

Human organoids were used to evaluate CAR-T efficacy in the relevant TME. A scRNAseq dataset of 161,970 cells was generated from ‘healthy’ and ‘myelofibrotic’ (TGFb/Gal1 rich) organoids co-engrafted with CD34+ myelofibrosis HSPCs and CAR-Ts. CAR-Ts exposed to mutCALR+ primary cells had a 2-fold expansion of CD8+ effector memory cells. The proportion of immunosuppressive, regulatory T cells (Tregs) increased in a high TGFβ environment, but not with Gal1. GSEA showed significant enrichment in pathways associated with IFNγ, TNFa and IL-2 in CAR-Ts exposed to mutCALR+ samples. Ongoing analyses will evaluate vulnerability of specific patient HSPC/lineage subsets to CAR-T, the ability of healthy (non-mutant) haematopoiesis to reconstitute after treatment, and the impact of CAR-T killing on the BM niche.

Conclusions:

Here we present a novel CAR-T therapy showing potent and remarkably selective targeting of mutCALR-driven malignancies. This study also evaluates the impact of JAKi on CAR-T efficacy and presents the first use of human organoids to evaluate immunotherapies in the relevant human tissue environment, including features of the TME, validating this as a powerful platform for pre-clinical development of targeted therapies across blood cancers.