Mutant calreticulin—Cancer driver and Achilles' heel for ET?

Essential thrombocythemia (ET) is a chronic blood cancer (Myeloproliferative Neoplasm, MPN) that arises following the acquisition of a mutation in a blood stem cell, causing enhanced signaling via the thrombopoietin receptor (TPO-R). Dr. William Dameshek was the first to speculate that ET, polycythemia vera (PV), and myelofibrosis were part of a unifying pathology, remarking in an editorial in 1951 that they were “perhaps due to a hithero undiscovered stimulus.”1 Half a century later, four teams published that gain-of-function mutations in the cytokine signal transducer JAK2 (JA2V617F) were present in almost all PV and 2/3 of ET and myelofibrosis cases.2-5 JAK inhibitors rapidly progressed through clinical development, and the first—ruxolitinib—was licensed for the treatment of myelofibrosis only 6 years after the identification of JAK2V617F. One year later, MPLW515L/K mutations were identified in 5% of patients with ET and myelofibrosis,6, 7 but the initiating driver mutation in the remaining 30% of JAK2/MPL mutation-negative cases remained enigmatic. In 2013, two studies published by independent teams in the same issue of the New England Journal of Medicine identified mutations in calreticulin (mutCALR) in 70%–80% of these patients, identifying the second most frequent MPN cancer driver mutation.8, 9 A role for CALR in MPN pathogenesis came as a surprise, given that it was known as an endoplasmic reticulum (ER)-resident chaperone protein and had not previously been implicated in cancer biology. The original publications described the most frequent mutations as a 52-base pair (bp) deletion (Type 1) or a 5-bp insertion (Type 2), but found that ~1 in 6 mutations were noncanonical. Crucially, all pathogenic mutations resulted in the same novel C-terminal epitope,8, 9 creating an aberrant, positively charged neoepitope with deletion of the ER-retention signal, “freeing” mutant CALR (mutCALR) from its normal intracellular localization. Subsequently, a series of publications confirmed the fascinating mechanism by which mutCALR triggers excess megakaryopoiesis, thrombocytosis, and myelofibrosis. First, the essentiality of the thrombopoietin receptor (TPO-R) was demonstrated.10 This was followed by a description of the sequence of events in which the positively charged neoepitope forms an aberrant complex with the TPO-R, and migration of the mutCALR-TPO-R complex to the cell membrane, triggering cytokine-independent receptor activation and downstream signaling (Figure 1). Fortuitously, from a therapeutic perspective, this also leads to a unique phenomenon in cancer biology: the direct product of a cancer-driving oncogene being displayed as a cell surface neoantigen, creating an opportunity for molecularly targeted, cancer cell-specific therapies. Notably, TPO-R expression is restricted to hematopoietic stem cells (HSCs), megakaryocytes, platelets, and their intermediate progenitors—and since mutCALR requires TPO-R for surface expression, the mutant protein is confined to these cells. As HSCs are the cancer-initiating cells and megakaryocytes are the key drivers of fibrosis, mutCALR-directed immunotherapies therefore primarily hit the cells responsible for both disease initiation and progression to advanced-phase MPN. ET results in a significant burden of morbidity and mortality. The incidence of mutCALR+ MPNs is estimated at 1–2.5/100,000 population per year, with a prevalence of 9–24/100,000 individuals.11 Over 50% of patients with ET experience fatigue or vasomotor symptoms such as headache, light-headedness, acral paresthesia, and erythromelalgia.12 Complications include thrombosis, bleeding events, progression to myelofibrosis, progression to leukemia, pregnancy complications, and immune compromise. Clear differences exist in the age of presentation and average platelet counts between JAK2V617F+ ET and mutCALR+ ET (Figure 2), with mutCALR+ cases being detected at a younger age (54 vs. 61 years for mutCALR vs. JAK2V617F+ ET) and with higher platelet counts (866 vs. 726 × 109/L).13 Notably, although JAK2, MPL, and CALR mutations all cause thrombocytosis via pathological activation of TPO-R signaling, their impact on platelet count and function appears distinct. In a review of over 1580 patients with MPN in the United States and Italy, the incidence of a platelet count >1000 × 109/L was 53% for Type 1 mutCALR+ ET and 71% for Type 2 mutCALR+ ET, compared to only 34% for JAK2V617F+ ET.14 Despite the higher platelet counts, the cumulative incidence of thrombosis is lower for mutCALR+ ET—at around 10% at 10 years—compared to 20% for JAK2V617F+ ET,8 an observation that is accounted for in the IPSET Thrombosis Risk Score by incorporating the presence or absence of the JAK2V617F driver mutation in the risk prediction algorithm.15 The combination of more extreme thrombosis with reduced thrombotic risk for mutCALR ET as compared to JAK2V617F+ ET likely reflects a negative impact of the mutation on platelet function (reduced adhesion, spreading, and response to agonists),16 as well as an increased likelihood of acquired von Willebrand disease with extreme thrombocytosis. Patients with Type 1 mutCALR+ ET have an overall better survival but face a paradoxically higher incidence of bone marrow fibrosis compared to those with Type 2 mutCALR or JAK2V617F+ ET.17 This is somewhat counterintuitive, given that leukemic transformation is the leading cause of death in MPN patients18 and that evolution to leukemia is substantially more common from myelofibrosis than from ET or PV. Therefore, while progression to myelofibrosis is a definite concern with mutCALR ET, the overall impact of this on survival and clinical outcomes remains unclear. The differences in clinical phenotypes between Type 1- and Type 2-mutCALR-driven MPNs are not completely understood but may relate to differences in loss of the calcium-binding function with Type 1 but not Type 2 mutations,19 as well as their impact on downstream signaling pathways. In a recent study, the mutations were shown to differentially activate unfolded protein response (UPR) pathways, with type 1 mutant cells being more dependent on BCL-xL for survival, potentially promoting TPO-R signaling-independent mutant cell survival.20 Although further research is required, it is possible that these features differentially impact megakaryocyte function and their capacity to drive fibrosis. In current management algorithms for ET, the decision to initiate therapy is largely based on the risk of thrombosis, rather than disease progression. The benefit of aspirin for primary prevention of thrombosis appears less certain in mutCALR+ ET than JAK2V617F+ MPNs, given the lower incidence of thrombosis and higher rates of extreme thrombocytosis and bleeding, and it may even confer a higher risk of bleeding without reducing thrombotic events.21 Consensus-based recommendations from the European LeukemiaNet suggest low-dose aspirin (75–100 mg/day) for patients with classical low-risk ET who have concomitant cardiovascular risk factors, but endorse observation only (without antiplatelet therapy) in patients without cardiovascular risk factors, and advise using antiplatelet agents with caution if platelet counts are >1000–1500 × 109/L due to the risk of bleeding.22 Indications for cytoreduction are also less clear-cut for mutCALR+ ET than for JAK2V617F+ MPNs. Non-controversial indications include a prior history of thrombosis and/or platelet counts over 1500 × 109/L, and those over the age of 60 who have significant cardiovascular risk factors.22 Whether patients over the age of 60 without cardiovascular risk factors also benefit from cytoreduction, and whether it is necessary to target a platelet count of 1000 × 109/L). Durable hematological responses were evident at lower doses in patients with Type I CALR mutations than those with non-Type 1 mutations. Most strikingly, nearly all patients (96%) achieved a reduction in mutCALR variant allele frequency (VAF), with 52% achieving reductions >25% and 31% achieving reductions >50%. Molecular responses were common alongside hematological responses.32 Encouraging efficacy was reported in the myelofibrosis cohort, with improvements in splenomegaly, symptoms, and anemia associated with molecular responses, although the tempo of clone size reduction appears slower in myelofibrosis than ET.33 The reasons for the slower molecular responses are unclear, and may relate to the greater burden of disease in myelofibrosis and higher abundance of mutCALR+ cell types that do not express the mutCALR-TPO-R on the cell surface, therefore are not directly vulnerable to mutCALR targeting therapies. Such cells would be expected to gradually reduce with time, if mutant HSCs are effectively targeted allowing healthy wildtype hematopoiesis to recover. Early translational analyses indicated clearance of mutant clone stem cells, megakaryocytes, and aberrant erythroid cells with normalization of bone marrow histology (reduced fibrosis and megakaryocyte clustering, increased proportion of non-mutCALR+ megakaryocytes, and increased bone marrow erythropoiesis).34 INCA033989 is Fc-silenced and therefore presumed to act independently of immune cell function. In contrast, JNJ-88549968 is a bispecific, T cell-engaging antibody (BiTE) that contains binding domains for both mutCALR and CD3, redirecting endogenous T cells to eliminate mutCALR+ cells.35 A phase 1 study is ongoing, in which a comparable safety profile to other T cell engagers would be predicted, with cytokine release syndrome expected to be among the therapy-emergent adverse events, reflecting on-target T cell activation. Given the encouraging data with INCA033989, this agent may also prove disease modifying. Other modalities at earlier stages of development include mutCALR-directed ADCs. Preclinical data were presented at the EHA and ASH 2025 congresses for ADCs with novel payloads, including SMARCA2/4 or CDK9 degraders.36 ADCs offer a compelling mechanism of action with targeted delivery of payloads directly to mutCALR-expressing cells, combining the specificity of antibody-targeting with cytotoxic potency of the payload, potentially offering more potent cytotoxicity than can be achieved with signaling inhibition or T-cell engagement in the context of T cell exhaustion. Internalization is required for ADCs to exert their mechanism of action, and payload-related toxicities will need to be considered. CAR T therapies targeting mutCALR are also in development,37-39 with preclinical evidence of selective cytotoxicity against patient cells in preclinical models, including human organoids and in vivo.37 The versatility and potency of CAR-T cell therapies are appealing, as T cells can be engineered not only to express novel receptors (overcoming the necessity of target peptide-HLA-expression for T cell engagement) but also to secrete payloads that might combat a challenging tumor microenvironment (TME). Effective clearance of target cells and long-lasting remissions have been demonstrated in other indications, but manufacturing costs and toxicities present ongoing challenges for CAR-T cell therapeutics. Overall, cell surface expression of mutCALR is intrinsically linked to MPL expression, and therefore varies across hematopoietic compartments, with higher levels on megakaryocyte lineage cells than HSCs. The low-level expression on cancer-initiating HSCs may hinder immunotherapies from achieving lasting molecular responses. However, CAR T cells can lyse target-antigen-expressing cells at antigen densities approximately 1000-fold lower than those required for antibody-mediated complement-dependent cytotoxicity. It is conceivable that a CAR-T cell therapeutic may enable superior targeting of mutant HSCs compared to a BiTE (albeit with increased cost and toxicities), although no direct comparisons have been published to date. CALR mutations, especially atypical variants, can result in structurally distinct C-terminal neoepitopes that have a differential impact on signaling. These differences may impact the binding and efficacy of therapeutic antibodies directed against the mutant C-terminus. In addition, increased TPO-R dimerization, implying greater signaling activation, has been reported for Type 2 (ins5) versus other variants, which may impede the impact of inhibitory antibodies, at least at lower doses.40 Achieving potent targeting of Type 2 variants is particularly important, given that patients with Type 2 CALRmut ET have higher platelet counts (albeit a lower incidence of fibrosis), and those with myelofibrosis have overall worse outcomes than those with Type 1 variants. Mutation selectivity may vary between antibodies, depending on the specific epitope targeted. To further overcome mutation-type selectivity, several distinct approaches have been proposed: (1) Targeting the conserved N-domain: INCA035784 is a T-cell redirecting antibody directed to the N-domain, a region that is invariant across known CALR mutations and also conserved with wildtype CALR, but only exposed for binding when the protein is complexed with the TPO-R. While no clinical data is available yet for this agent, targeting common regions of the C-terminus or the N-domain may be one way to overcome mutation-type selectivity41; (2) Combining antibodies that target distinct epitopes on the C-terminus: A second approach proposed to enhance antibody-mediated signaling inhibition is to combine antibodies targeting distinct domains on the C-terminus. A recent report combining proximal- and distal-epitope binding antibodies demonstrated significantly enhanced signaling inhibition for Type 1 and Type 2 variants. While antibody combinations alone did not achieve complete inhibition for Type 2 CALR mutations, efficacy further synergy was demonstrated by combining dual antibody targeting with ruxolitinib, suggesting that patients with Type 2 variants may benefit from triple therapy with dual antibody targeting and JAK inhibition.40 (3) Exploiting differential impact on signaling: Differential signaling between Type 1 and Type 2 mutations are a potential therapeutic vulnerability, with differential dependence on BCL-2 versus BCL-xL targeting that might inform selection of targeted The and of response reported for INCA033989 is that mutCALR targeting therapies may the first MPN However, the key First, the of mutCALR cell surface expression is linked to expression of at low and even lower on HSCs than on cells. Whether mutCALR expression on HSCs is to of the cancer-initiating and cells remains This will only be evident when patients with molecular responses after treatment the of mutant clone Similarly, the of molecular response is not MPNs the driver mutations acquired several to clinical It is therefore possible that a reduction in of may the to substantially clinical outcomes for patients, and that molecular responses may not be treatment for However, if the cell is targeted, the mutant clone burden may rapidly following of therapy. the in a for In ET and myelofibrosis, bone marrow megakaryocytes and studies have associated T cell in mutCALR Whether T immune therapies can overcome these in all patients and if the fibrosis will stem cells from targeting remains to be If inhibitory antibodies can achieve disease that to immune but not molecular remissions for patients, it is possible that targeting approaches may be with monoclonal antibody treatment followed by an BiTE or CAR-T cell therapy. Although in the of patients, mutations in the same clone as the MPN driver they can also in This can when the mutation the MPN driver and a clone, or if an mutation in a in a patient with an mutCALR or inhibition of the mutCALR+ clone in these cases will and may of but is that the may be more than the previously mutCALR If of the at cell may be required prior to treatment for patients with of This concern is also a to treatment earlier in the disease prior to the acquisition of if the clinical and molecular responses observed in the early studies of targeted antibodies prove durable and are associated with improvements in longer-term a of responses to therapy are for patients with mutCALR+ ET is for in the as well as for as clinical and therapies an opportunity to the management of MPNs. A of mutCALR-directed therapies are and a of targeting modalities is likely to be required, as distinct and data of current treatment and In a future of therapies and MPNs based on the driver mutation may be more than to cell types or clinical This is particularly for ET, and such distinct and clinical Similarly, of and therapies that disease will present a compelling for younger patients with is, treatment algorithms from on thrombotic to a approach (Figure the risk ET patients can from a significant symptom burden that as well as an increased risk of thrombosis, pregnancy complications, and This the and clinical for treatment and the of treatment in ET from events to disease As the first of achieving reduction in time, the also data to inform the of molecular therapy even the of and and is a and of has research from and and for and/or from and has no This was by by in with for and for is not to this as no were or the current