CAR-T Targeting of Mutant Calreticulin Establishes a Potentially Curative Stem Cell-Directed Therapy for Myeloproliferative Neoplasms

Myeloproliferative neoplasms (MPNs) are sustained by mutated hematopoietic stem cells (HSCs). Existing therapies fail to eliminate this compartment, leaving allogeneic HSC transplantation as the only curative option. Recurrent MPN driver mutations in calreticulin (CALRmut) generate a C-terminal neopeptide that requires cell-surface expression for oncogenic signaling, making it an attractive immunologic target. However, it remains unknown if CALRmut is uniformly displayed on all MPN HSCs within hematopoietic microenvironments. We generated huAB2, a high-affinity CALRmut-specific humanized antibody, to use as the targeting domain for chimeric antigen receptor (CAR)-T cells. We show that CALRmut is consistently displayed on functional MPN HSCs and accessible in vivo. huAB2 CAR-T cells eradicate MPN-propagating CALRmut HSCs in patient-derived tumor xenograft models without antigen escape while preserving coexisting normal human and host hematopoiesis. These findings establish CALRmut display as an obligate feature of MPN HSC fitness and support the feasibility of curative, non-transplant immunotherapy for CALRmut MPNs. SignificanceTherapies that eradicate cancer stem cells enable cure, but their feasibility is unknown. We establish an approach to potentially cure MPNs by proving mutant calreticulin to be a MPN stem cell marker that can be targeted by CAR-T cells to selectively wipe out disease in preclinical models of human MPNs.