In the study by Tirrò et al. , the authors describe a patient presenting two myeloproliferative neoplasms (MPNs) and harboring a JAK2 V617F mutation and a BCR-ABL1 fusion transcript.
MPNs are clonal hematological diseases, which are chronic but can progress to leukemia. They include several entities such as chronic myeloid leukemia (CML), classical MPNs with essential thrombocythemia (ET), polycythemia vera (PV), and primary myelofibrosis (PMF), as well as much rarer entities such as mastocytosis, chronic eosinophilic leukemia, chronic neutrophilic leukemia, and other unclassifiable MPNs. They are all due to genetic abnormalities occurring at the level of the hematopoietic stem cell that induce a constitutive tyrosine kinase activity. CML is the archetype of these diseases and is characterized by an overproduction of granulocytes. This phenotype is due to the translocation t(9,22) corresponding to the Philadelphia chromosome and resulting in the chimeric protein BCR-ABL1. Classical MPNs lead to an overproduction of red blood cells in PV, platelets in ET, and to a deregulation of megakaryocytic and granulocytic lineages in PMF. They are due to mutations in genes that overactivate the JAK2 kinase. Mutations in JAK2 (JAK2 V617F) are detected in around 50% of the ET and more than 95% of the PV, and mutations in JAK2 exon 12 are only found in PV. Mutations in the calreticulin gene (CALR m) are only found in 25–30% of the ET and PMF cases, and the mutant CALR constitutively activates JAK2 through binding to the thrombopoietin receptor (MPL). Moreover, 5–10% of the ET and PMF cases display mutations in the gene encoding MPL(MPL W515K/R/A/K/L) .
BCR-ABL1 and non-BCR-ABL1 MPNs are rare with an incidence of 1/100,000 cases, and the frequency of co-occurrence of two disorders in the same patient is even rarer . Tirrò et al.  describe a patient presenting with ET at diagnosis and then 3 years later with CML and harboring both the JAK2 V617F mutation and a BCR-ABL1 fusion transcript. Such cases were reported with a 0.2–2.5% recurrence in large MPN cohorts. Virtually, all combinations of mutations can exist, giving rise to several diseases in the same patient underlying the great heterogeneity of MPNs. The resulting diseases are then driven by each of the different phenotypic mutations. Some of these co-occurrences have been described and include not only BCR-ABL1/JAK2 V617F (CML and PV or ET or PMF); BCR-ABL1/CALR m (CML and ET or PMF); BCR-ABL1/JAK2 exon 12 (CML and erythrocytosis) but also CALR m/JAK2 V617F and MPL W515R/CALR m harboring two classical MPNs (Fig. 1).
Interestingly, in the present study, the authors used colony assays to determine the clonal architecture of the genetic abnormalities in the progenitors. This strategy permits one to define if they are in the same clone or not and thus if the diseases are independent or not. The authors describe that BCR-ABL1 arises in a JAK2 V617F progenitor, as reported by several other groups that have described CML with an ET, PV, or PMF. Moreover, the same clonal structure was identified with CALR m in a patient with CML and ET . Furthermore, no study has formally established that BCR-ABL1 and JAK2 V617F mutations could be in two independent clones. In contrast, when CALR m and JAK2 V617F were found together (4.2% of the ET cases), they were in independent clones and remained exclusive. Most of the time and as described in this study, the BCR-ABL1/JAK2 V617F clone outcompetes JAK2 V617F clone alone, showing its high clonal dominance and the dominance of CML over ET. Similarly, when found in the same patient, the CALR m clone outcompetes the JAK2 V617F clone, which remains very low (<20% of the cells) .
The mechanism of acquisition of several phenotypic drivers in the same patient is still unclear. In fact, the frequency of each individual mutation is rare, so their association would be so unlikely that it is expected that no patient or only extremely few will be found. To explain this paradox, several hypotheses can be postulated. First, a genetic predisposition may exist that can favor either (i) the genetic instability and the generation of several mutations in different or similar clones; (ii) the selection of preexisting mutated progenitor cells; or (iii) both. Few predisposing factors have been depicted in classical MPNs. Among them are the 46/1 haplotype, TERT variant and RBBP6 mutations, which deregulate the P53 pathway and induce a potential genetic instability. Moreover, the ATG2B/GSKIP locus can predispose to all hematological malignancies, sometimes including CML and ET in the same patient harboring CALR m and BCR-ABL1 . Hematopoietic aging or other acquired associated mutations such as TET2 can also lead to a predisposition to acquire driver mutations. Alternatively, since BCR-ABL1 appears after JAK2 V617F in this study and others, it could also be hypothesized that JAK2 V617F-mediated replicative stress and increased homologous recombination in progenitors could favor genomic translocation such as BCR-ABL1 . Another hypothesis could be that JAK2 V617F mutations or BCR-ABL1 translocation could be much more frequent than we think, but that the initiation of a disease by these mutations is a rare event. This hypothesis is supported by the detection of JAK2 V617F mutations or BCR-ABL1 in the normal population. Therefore, when the disease is already initiated by JAK2 V617F, the probability to get BCR-ABL1 remains the same as in the normal hematopoiesis.
The study by Tirrò et al.  also reports the genetic complexity of this case with a branched structure. Indeed, they show evidence indicating the loss of heterozygosity of JAK2 V617F in two independent clones and the detection of homozygous JAK2 V617F clones with and without BCR-ABL1. This genetic complexity has also been shown in other cases with multiple clones. Moreover, several patients with BCR-ABL1 and JAK2 V617F can harbor other associated mutations in molecules affecting epigenetics (TET2, ASXL1, and DNMT3A in around 50%), controlling spliceosome (SF3B1), and in transcriptional factors (P53, RUNX1, PHF6) . Those transcriptional factors, in addition to JAK2 V617F and BCR-ABL1, could also be involved in the increased genetic instability, resulting in the complex genetic structure of these patients.
In the study discussed here , the authors used first- and second-generation tyrosine kinase inhibitors (TKIs; nilotinib, dasatinib) to achieve the complete cytogenetic response for BCR-ABL1, probably without affecting the JAK2 V617F clone. In this regard, deciphering the order of mutations is of interest, since TKIs will generally target the BCR-ABL1 clone but not JAK2 V617F alone, although the second-generation TKIs such as dasatinib were suggested to target both. Novel tailored care and treatment strategies for such patients could be used, such as the combination of TKIs with IFNα, which presents hematological and molecular efficacy both in CML and ET. Alternatively, combining TKIs with ruxolitinib (JAK2 inhibitor) could also be considered in PV or PMF stages for its clinical efficacy.
The author thanks Caroline Marty and William Vainchenker for the critical reading of the manuscript.
The author has no conflicts of interest to declare.
This work was supported by INSERM and grants from INCA-PLBio 2015 and 2017, and Labex GR-Ex is funded by the program “Investissements d’avenir.”