Paper Details

ABSTRACT

Background: Myeloproliferative neoplasms (MPNs) are chronic cancers that cause significant morbidity and shorten survival. Although current therapies improve symptoms and some disease manifestations, they do not eradicate the mutated hematopoietic stem cells (HSCs) that initiate and sustain disease. Allogeneic hematopoietic stem cell transplantation remains the only curative therapy but is suitable for only a minority of patients. Over 30% of MPNs harbor frameshift mutations in calreticulin (CALR) that generate a shared neoantigen displayed with the thrombopoietin receptor (MPL) on the surface of MPN HSCs, providing an attractive target for immunotherapy. Here, we developed mutant calreticulin (CALRmut)-directed chimeric antigen receptor (CAR)-T cells and evaluated whether they can selectively eradicate primary patient-derived CALRmut MPN HSCs in vivo.

Methods:  We established a translational platform for CALRmut-targeted immunotherapy encompassing antibody discovery, target validation on HSCs, CAR-T cell engineering and progressively rigorous disease models, culminating in a novel patient-derived tumor xenograft (PDTX) platform supporting robust engraftment of primary human CALRmut MPN HSCs for in vivo therapeutic testing.

After screening over 7,000 monoclonal antibodies, we identified a humanized antibody (AB2) which binds the shared CALRmut neopeptide with sub-nanomolar affinity without detectable binding to wild-type CALR (CALRwt). Target engagement on MPN HSCs was confirmed in vivo by selective binding of fluorescently-labeled AB2 to CALRmut HSCs in syngeneic knock-in mice expressing human CALRmut.

Using the AB2 binding domain, we generated CALRmut-directed AB2 CAR-T cells. Donor-matched CD19-directed CAR-T cells (CD19 CAR-T) and untransduced T cells (UTD) were used as negative controls. Activity was evaluated using engineered Ba/F3 cell line models expressing human MPL with type 1 or type 2 CALRmut, primary patient-derived CALRmut MPN hematopoietic stem and progenitor cells (HSPCs), and our new CALRmut MPN PDTX platforms. Non-target controls included Ba/F3 cells expressing human MPL with CALRwt, healthy donor (HD)-derived CALRwt HSPCs, and patient-derived JAK2-mutated MPN HSPCs.

Results: AB2 CAR-T cells selectively eliminated Ba/F3 cells expressing either type 1 or type 2 CALRmut, with no activity against CALRwt cells. In a highly aggressive Ba/F3 xenograft mouse model that uniformly killed untreated and control-treated mice within 2 weeks, AB2 CAR-T cells eradicated disease, markedly prolonged survival, and achieved durable cure in 40% of treated mice. AB2 CAR-T cells also selectively eliminated primary patient-derived CALRmut HSPCs in vitro while sparing HD and JAK2-mutated MPN negative controls. Selectivity was further demonstrated by unbiased spectral immune profiling and multiplex secretome analysis that showed that AB2 CAR-T cell activation, cytokine release, and memory subset expansion occurred only in response to CALRmut target cells.

In our newly developed CALRmut MPN PDTX model, AB2 CAR-T treatment eliminated CALRmut hematopoiesis in vivo, whereas UTD- and CD19 CAR-T-treated mice showed progressive marrow replacement by CALRmut cells and died from disease. Following elimination of the CALRmut clone, AB2 CAR-T treatment enabled selective outgrowth in vivo of residual CALRwt human HSCs that were barely detectable in the donor grafts. In chimeric PDTX mice receiving mixed male CALRmut/female HD grafts, AB2 CAR-T treatment selectively eliminated the male CALRmut clone while preserving female CALRwt hematopoiesis, confirming selective elimination of CALRmut HSCs in vivo.

Critically, functional eradication of CALRmut MPN HSCs from AB2 CAR-T-treated PDTX mice was demonstrated by the absence of CALRmut alleles from sorted HSCs and colony-forming cell clones and by the absence of CALRmut engraftment after serial transplantation. With male CALRmut/female HD chimeric grafts, bone marrow from AB2 CAR-T-treated mice reconstituted exclusively female CALRwt hematopoiesis, whereas bone marrow from control-treated mice re-established male CALRmut disease. AB2 CAR-T cells thus selectively eradicate functional MPN HSCs.

Conclusions: To our knowledge, these studies provide the first preclinical proof-of-concept that selective immunologic eradication of human MPN HSCs while preserving normal HSCs is achievable in vivo, establishing a foundation for transplant-free curative therapies for CALRmut MPNs.