Mutations in MPNs to “interfere-on”

The study conducted by Mosca et al¹ in this issue of Blood details the analysis of serial blood samples from 48 patients with chronic myeloproliferative neoplasms (MPNs) treated with interferon α (IFNα) over a 5-year period. This elegant work examines the impact of IFNα therapy on the clonal dynamics of malignant hematopoietic stem and progenitor cells (HSPCs) and identifies driver mutation type, JAK2V6217F homozygosity, and IFNα dose as independent predictors of therapeutic response. Specifically, these results indicate that progenitor cells harboring homozygous JAK2V617F were more susceptible to lower doses of IFNα than heterozygous cells, and CALR type 2 mutant-bearing progenitors were more effectively targeted than type 1–positive cells. Using mathematical modeling and a statistical inference method, the authors concluded that IFNα was able to exhaust the JAK2V617F hematopoietic stem cell (HSC) pool through induction of differentiation. In the case of mutant CALR, this effect was preferentially seen in type 2 mutants. Additionally, IFNα was more effective in activating (ie, exiting from quiescence) JAK2V617F homozygous over heterozygous HSCs. These studies dissect the differential effects of IFNα on the hematopoietic cell compartments and highlight the influence of MPN driver mutation on clonal dynamics.

IFNα has been used for decades to treat hematologic malignancies and, before the advent of tyrosine kinase inhibitors, was the agent of choice to induce hematologic, cytogenetic, and even molecular responses in a subset of patients with BCR-ABL1–positive MPN. Numerous studies have revealed the pleiotropic effects of IFNα on the malignant cell population, including enhanced immune response, induction of apoptosis, inhibition of angiogenesis, and exiting from cellular quiescence, which are expertly reviewed elsewhere.² 

Whether the molecular observations of Mosca et al can be effectively integrated into routine care of patients with MPN in determining appropriate patient selection and dosing of IFNα based on genotype requires validation in a prospective trial. For example, should patients with essential thrombocythemia (ET) and polycythemia vera (PV) harboring JAK2V617F be treated with a continued upward titration of IFNα to a maximally tolerated dose to optimize the effect on the MPN HSCs and maximize the opportunity to attain a molecular response? Conversely, should ET patients harboring CALR type 1 mutation be maintained on lower doses with a different expectation of molecular response kinetics? These types of treatment decisions would also need to be balanced with the well-described grade 3/4 treatment toxicities and quality-of-life impact that can be associated with prolonged IFNα therapy (see figure).

It is important to emphasize that the current primary goal of therapeutic intervention in ET and PV is to reduce the incidence of thrombohemorrhagic events. The ET/PV-based risk stratification systems used in clinical practice identify those patients at increased risk of thrombosis. Cytoreduction with hydroxyurea (HU) or IFNα are considered first-line treatment options for PV and, additionally, anagrelide can be used for ET. Recently published randomized phase 3 clinical trials (PROUD/CONTINUATION-PV and Myeloproliferative Disorders–Research Consortium 112) comparing HU to IFNα found superior hematologic and molecular responses with IFNα over extended treatment periods (ie, after 2 years of therapy).^3,4 However, neither trial was conducted for the purpose of nor is capable of confirming superiority of reduction in the rate of thrombosis or freedom from progressive disease. The primary end point of these and similar trials is achievement of complete hematologic response using European LeukemiaNet criteria. However, this end point driven by blood count normalization, elimination of splenomegaly, and alleviation of symptom burden is not a proven surrogate for either prevention of thrombosis or progression-free survival.

The current aim of translational research efforts in MPNs is the effective targeting and deletion of the malignant HSCs while concurrently dismantling the tumor-supporting microenvironment to obtain histomorphologic, cytogenetic, and molecular remission. The benefits of such responses at this point are still inferred but presumably include disease course modification and prolongation of survival. Recent reports of novel mechanism-based agents such as MDM2 antagonists and mutant CALR vaccine approaches offer intriguing combination options with IFNα.^5,6 The future of MPN therapy in general includes earlier therapeutic intervention and incorporation of multiagent regimens such as combination IFNα and JAK inhibitor therapy in PV.⁷ 

Currently, driver mutation status and variant allele fraction do not dictate treatment decisions in MPNs with IFNα or, for that matter, any cytoreductive agent. In contrast to the molecular observations found in this report, we found that CALR mutation status in ET was associated with a higher complete hematologic response rate to pegylated IFNα2a in the MPDRC 111 trial.⁸ Interestingly, recent data in myelofibrosis (MF) would appear to support the concept of differential responses with JAK2 inhibitor therapy based on driver mutant allele burden status in patients with MF treated with pacritinib compared with ruxolitinib.⁹ Given the molecular overlap across the MPN disease spectrum, one could imagine a molecularly defined treatment approach where type and dose of driver mutation and presence of non–JAK-STAT pathway involved mutations (epigenetic, spliceosome, etc)¹⁰ influence treatment selection and delivery. The sequence of MPN mutation acquisition may also impact the biologic response to IFNα therapy. Additionally, the potential of IFNα discontinuation after attainment of a molecular response may allow a subset of patients with MPN to remain in a treatment-free remission. However, until prospective validation of molecular determinants of response and treatment-free remission are completed, IFNα should be used in the appropriate clinical settings irrespective of MPN genotype and with the primary intention of thrombosis reduction. The promise of disease course modification, even in the presence of molecular response, has not yet been confirmed.