Sequential Development of JAK2^V617F Mutation and BCR-ABL1 Fusion in Individual Patients With Myeloproliferative Neoplasms: A Linear Clonal Evolution or Parallel Clonal Competition?

Six cases of MPN with concomitant JAK2^V617F mutation and BCR-ABL1 fusion were identified from our bone marrow biopsy database. These included 4 cases from Duke University Medical Center (Durham, North Carolina) and 2 cases from the University of Washington (Seattle). The diagnosis of primary myeloproliferative neoplasm in each case was confirmed according to the 2017 update of WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues.^1–3  The clinical history and laboratory data for the patients were collected from corresponding clinical notes in our medical informatics systems.

Peripheral blood smears were stained with Wright stain, bone marrow aspirate smears and biopsy touch imprints were stained with Wright-Giemsa stain, and bone marrow core biopsies and clot sections were stained with hematoxylin-eosin stain. The cases were reviewed independently by 2 hematopathologists (E.W. and D.R.). Bone marrow cellularity was evaluated on core biopsy, or on clot section in cases where the biopsy quality was suboptimal. The ratio of myeloid to erythroid elements was calculated from differential cell counts on bone marrow aspirate smear or touch imprint, or estimated on the histologic section of either bone marrow biopsy or clot, if the cytomorphologic sample was inadequate. The size of megakaryocytes, their nuclear lobation, and their distribution were evaluated on biopsy and/or clot sections.

Cytogenetic analysis was performed on 2 to 4 mL of bone marrow aspirate from each case. Cultures from each specimen were initiated from the fresh, anticoagulated specimen in tissue culture medium. The cells were incubated for 24 and 48 hours without mitogen stimulation. Chromosome preparations, including harvesting and GTW banding, were made using standard methods. Cytogenetic abnormalities were classified according to the International System for Human Cytogenetic Nomenclature.

An interphase fluorescence in situ hybridization (FISH) was performed using a Vysis dual-color, dual-fusion BCR-ABL1 rearrangement probe. This probe is designed to detect the juxtaposition of the BCR locus and ABL1 gene sequences. The translocation involving BCR at 22q11.2 and ABL1 at 9q34, t(9;22)(q34;q11.2) is visible by the fusion of the red color probe and green color probe. In total, 200 interphase nuclei were evaluated by 2 different technologists, and the percentage of positive cells was reported.

This assay used the QuantideX minor and IS kits (Asuragen Inc, Austin, Texas) to quantitate BCR-ABL1 mRNA in peripheral blood or bone marrow using endogenous ABL1 transcripts as reference. The lower limit of detection for the minor kit was set at 0.0025% cells expressing BCR-ABL1 fusion transcripts, whereas the lower limit of detection for the IS kit was set at 0.002% cells expressing BCR-ABL1 fusion transcripts. Results for the p210 transcripts were reported on the International Scale.

JAK2^V617F mutation was tested by 3 methods because of the chronological advance in detection techniques. These included qualitative allele-specific polymerase chain reaction (PCR) assay, semiquantitative PCR assay with allele-specific primers, and next-generation sequencing analysis of myeloid panel. All these assays used high-quality genomic DNA prepared from peripheral blood or bone marrow aspirate sample.

This assay used allele-specific primers targeting both wild-type JAK2 and mutant JAK2 alleles. After PCR amplification, fluorescently labeled products were separated by capillary electrophoresis. The presence of both PCR products, or the presence of only correctly sized mutant product, indicated that cells harboring JAK2^V617F mutation were present in the sample. The assay had a detection limit of 0.1% in a background of wild-type genomes.

This assay used a TaqMan real-time PCR-based technique to detect the V617F mutation (c.1849G>T) in the JAK2 gene. A single real-time PCR reaction was performed in each assay using an oligonucleotide primer pair that amplifies both the mutant and wild-type JAK2 alleles and allele-specific TaqMan probes that were fluorescently labeled with VICTAMRA (wild-type JAK2) or FAM-TAMRA (V617F mutant JAK2). A threshold cycle value (Ct) was measured for each TaqMan probe. For each sample, a δ Ct value was calculated as the difference in Ct value between the JAK2 wild-type and JAK2^V617F mutant probes. In addition, 1%, 50%, and 100% mutant controls, as well as wild-type control, were included in each assay. Results were interpreted based on the patient δ Ct value relative to the controls. This test was performed using an ABI Prism 7500 Sequence Detection System (Thermo Fisher Scientific Inc, Waltham, Massachusetts).

Specific regions of 75 target genes (including the hotspot mutation locus of the JAK2 gene) were sequenced using the Illumina NextSeq 500 (Illumina Inc, San Diego, California). The resulting DNA sequence was assembled and aligned to the reference human genome version GRCh37/hg19. Sequence analysis was performed using the Archer Analysis bioinformatics pipeline. The lower limit of detection for this assay was 5%, meaning that a variant allele can be detected in a 95% wild-type allele background (or 10% neoplastic cell content). A depth of coverage of less than 100× was considered inadequate.

The clinical findings, pathologic features, and genomic evolution of MPN in 6 cases are summarized in the Table.

Of 6 patients, 4 were male and 2 were female. Age ranged from 46 to 77 years, with a median of 54 years at the time of initial presentation when diagnosis of primary MPN was made. Two of the patients had a prior history of regional radiotherapy for nonhematolymphoid neoplasms before their diagnosis of MPN (patients 4 and 6). In terms of the chronologic sequence of the 2 MPNs in individual patients, 5 patients (patients 1–5) received an initial diagnosis of JAK2^V617F-positive MPN, and 1 patient (patient 6) received a diagnosis of BCR-ABL1⁺ CML before developing JAK2^V617F-positive primary myelofibrosis 1 year later. The 2 patients with polycythemia vera were initially treated with phlebotomy, followed by hydroxyurea, and both received additional treatment with ruxolitinib after developing splenomegaly and anemia. The other 3 patients with other JAK2^V617F-positive MPN were initially treated with hydroxyurea (patients 3 and 4) or ruxolitinib (patient 5) before developing BCR-ABL1 fusion. In those 5 cases of JAK2^V617F-positive MPN, the latency from diagnosis of JAK2^V617F-positive MPN to acquisition of BCR-ABL1 fusion ranged from 4 to 13 years, with a median interval of 9 years. Patient 6 received nilotinib for the treatment of CML before development of splenomegaly and subsequent diagnosis of JAK2^V617F-positive myelofibrosis.

All 5 patients with an initial diagnosis of JAK2^V617F-positive MPN showed morphologic features of bone marrow biopsies consistent with the diagnosis of each patient given in the Table. Specifically, the biopsies in patients 1 and 2 exhibited hyperplastic hematopoiesis with proliferation of large lobate megakaryocytes in keeping with the diagnosis of polycythemia vera (Figure 1, A). Bone marrow biopsy of patient 3 showed slightly increased to normal cellularity with hyperplasia of large lobate megakaryocytes in line with the diagnosis of essential thrombocythemia (Figure 1, B). The biopsies in patients 4 and 5 demonstrated hypercellular bone marrow with panmyelosis including megakaryocytic hyperplasia (data not shown). Although significant fibrosis was not identified at the time, both patients developed severe myelofibrosis in subsequent biopsies, fulfilling the diagnosis of primary myelofibrosis. At the time of acquiring BCR-ABL1 fusion, all 5 patients with an initial diagnosis of JAK2^V617F-positive MPN demonstrated marked leukocytosis with left-shifted neutrophilia, and 4 of them exhibited leukoerythroblastosis. Patient 6 presented with left-shifted neutrophilia and basophilia at the initial diagnosis of CML and showed anemia and thrombocytopenia when a bone marrow biopsy identified JAK2^V617F-positive myelofibrosis. All 5 patients with secondary BCR-ABL1 fusion showed hypercellular bone marrow with myeloid hyperplasia and megakaryocytic hyperplasia (Figure 2). Of note, all 5 of these patients demonstrated hypolobate megakaryocytic nuclei, and all but 1 (patient 1) displayed relatively small megakaryocytes. The bone marrow biopsy in patient 1 was remarkable for large megakaryocytes with hypolobate nuclei, in contrast to the small megakaryocytes with hypolobate nuclei in the other 4. A total of 5 out of 6 patients demonstrated significant myelofibrosis, including patient 6, with severe myelofibrosis and osteosclerosis.

Of the 5 patients with initial JAK2^V617F-positive MPN, 2 patients (patients 1 and 4) had BCR-ABL1 fusion detected by chromosomal analysis, interphase fluorescence in situ hybridization (FISH) study, and quantitative reverse transcriptase–PCR, 1 patient (patient 3) by chromosomal analysis and reverse transcriptase–PCR, 1 patient (patient 5) by chromosomal analysis alone, and the remaining patient (patient 2) by quantitative reverse transcriptase–PCR alone. At the time of acquiring BCR-ABL1 fusion, 4 patients (patients 1–4) remained positive for JAK2^V617F, and the remaining patient (patient 5) did not have the mutation tested. Of the patients with retention of JAK2^V617F, 2 (patients 1 and 2) showed JAK2^V617F allele frequency higher than 50%, 1 showed it lower than 20% (patient 3), and the remaining patient (patient 4) had JAK2^V617F tested by qualitative allele-specific PCR assay, thus providing no quantitative data for the mutant allele. Of note, patients 1 and 2 demonstrated high proportions of cells harboring BCR-ABL1 fusion while retaining high allele frequencies of JAK2^V617F mutation. Patient 6 initially received tyrosine kinase inhibitor (TKI) therapy, achieving a 4-log reduction of BCR-ABL1 transcripts in 6 months, but developed splenomegaly, anemia, and thrombocytopenia approximately 1 year after the diagnosis of CML. Bone marrow examination demonstrated myelofibrosis with evidence of acquired JAK2^V617Fmutation but a complete loss of BCR-ABL1 by interphase FISH and quantitative reverse transcriptase–PCR testing.

With regard to subsequent treatment after acquiring additional genomic alterations, 4 patients (patients 3–6) received TKI plus JAK kinase inhibitors and/or hydroxyurea. Patient 2 also received TKI but died of disease complications 4 months after acquiring BCR-ABL1 fusion. Patient 1 was treated with radiotherapy targeting his enlarged spleen, but his peripheral blood smear and bone marrow biopsy remained similar in morphologic features 4 months later, with parallel increases in the quantities of BCR-ABL1 and JAK2^V617F allele burdens (Table, treatment and disease course). Patient 3 was treated with combined TKI and hydroxyurea, and his complete blood cell counts stayed within a relatively normal range, but the level of BCR-ABL1 transcripts fluctuated between <1% and 48% with periodic dose adjustment of TKI therapy. Patient 6 was treated with ruxolitinib, hydroxyurea, TKI, and supportive therapy and remained in a poor hematologic condition 26 months after diagnosis of JAK2^V617F-positive MPN. In patients 1, 4, and 5, the patients eventually received hematopoietic stem cell transplants and were alive with complete donor cell engraftment and without detectable JAK2^V617F or quantifiable BCR-ABL1 4, 49, and 17 months after transplantation, respectively.

Both JAK2^V617F mutation and BCR-ABL1 fusion are drivers leading to MPN, with a subtle difference in morphology and clinical manifestation between the 2 genomic alterations. The 2 genomic events were previously thought to be mutually exclusive, with only rare cases of both events concomitant in individual patients. Recently, Soderquist et al¹⁶  reported 11 cases of MPN with concurrent JAK2^V617F mutation and BCR-ABL1 fusion, estimating a prevalence rate of 0.4% among all MPNs tested for both genomic alterations. Through analysis of longitudinal samples in each individual case, they concluded that most instances of concurrent JAK2^V617F mutation and BCR-ABL1 fusion were actually composites of CML and BCR-ABL1⁻ MPN, and in some cases the second MPN was initially cryptic and became unmasked only after TKI treatment. This phenomenon of biclonal myeloproliferative processes (Figure 3, combination of routes 1 and 2) has been described in many case reports,^{9,10,12–15,17}  and it seems to explain the pathologic evolution in patient 6 and possibly in patients 3 to 5 of our series. In patient 6, the BCR-ABL1⁻ MPN emerged only after CML declined to an undetectable level status after nilotinib therapy, reflecting an impact of clonal competition between the 2 MPNs. Although there was no sample intermediate between CML and BCR-ABL1⁻ MPN to be tested for JAK2^V617F, which would thus demonstrate transitional change, the short interval between the 2 MPNs suggests the possibility of their composite during the transition or even before the initial diagnosis. A composite of the 2 neoplastic components is suggested by the cytogenetic study and molecular data for BCR-ABL1 and JAK2^V617F in patient 3. When BCR-ABL1⁺ CML is present in conjunction with JAK2^V617F-positive MPN, it seems to play a dominant role compared with the JAK2^V617F clone, quickly replacing the latter before TKI treatment,^15,16  as seen in patients 4 and 5 in our series. Many times, effective treatment with TKI reverses the relative prevalence of the 2 clones.^13–16 

Conversely, patients 1 and 2 in our series demonstrated a high allele frequency of JAK2^V617F when BCR-ABL1 fusion emerged with a high fraction of the fusion gene and/or fusion transcripts, suggesting a probable linear clonal evolution by acquiring this second genomic aberration on the existing JAK2^V617F-positive clone (Figure 3, combination of routes 1 and 3). This is especially evident in patient 1, in whom chromosomal analysis revealed 2 separate clones, one with isolated t(9;22) and the other with isolated 20q deletion. It is possible that both clones harbored JAK2^V617F mutation as well, given its high allele frequency (84%). With time, the fraction of Philadelphia chromosome and BCR-ABL1 fusion transcripts increased, whereas JAK2^V617F remained high in allele burden, suggesting an expansion of JAK2^V617F/BCR-ABL1 double-positive subclone (“double-hit” clone; route 3 in Figure 3).¹⁸  A JAK2^V617F allele frequency approaching 100% may be explained by homozygous mutation or loss of heterozygosity of the JAK2 locus in this case. Of note, the previous investigations relied on the dynamic change of allele burdens between JAK2^V617F and BCR-ABL1 in response to TKI to determine the clonal composition. This analysis may overlook or underestimate the possibility of clonal evolution to a “double-hit” MPN because a case with a “double-hit” subclone might respond to TKI treatment with a decline in BCR-ABL1 burden but persistence of JAK2^V617F, resembling that of a composite MPN.^13–17  On the other hand, a JAK2^V617F-positive subclone evolved from BCR-ABL1⁺ CML (Figure 3, composition of routes 2 and 4), a reverse sequence of “double-hit,” has not yet been well described either at the clinical or genomic level, even though it is theoretically feasible.¹⁴  This may be explained by the sustained minimal level of the BCR-ABL1⁺ clone in many patients due to effective TKI treatment, thus lowering the chance for a “second hit,” when CML occurs before the JAK2^V617F mutation. Given the potential clinical application, we advocate the genomic analysis of chronologic bone marrow biopsies to provide a mutational landscape and thus determine the genealogic tree in leukemia evolution.

The latency from JAK2^V617F-positive MPN to BCR-ABL1 acquisition took 4 to 13 years with a median of 9 years in our series, which is similar to the median interval reported by Soderquist et al¹⁶  and others.¹⁴  In addition, this latency is reminiscent of the interval between MPN and its transformation to acute-phase or myelodysplastic syndrome.^19,20  Of note, 4 of the 5 patients in our series were treated with hydroxyurea for their initial MPN. Although there is believed to be a low risk of mutagenesis with therapeutic doses of hydroxyurea,²⁰  the rarity of secondary BCR-ABL1 in patients with primary MPN suggests acquisition of this fusion event via a natural course rather than an iatrogenic cause. Interestingly, the clinical presentation and morphology seem to correlate with clonal composition, with onset of leukocytosis and small hypolobate megakaryocytes observed when acquiring BCR-ABL1 fusion, and thrombocytosis and large lobate megakaryocytes seen after acquisition of JAK2^V617F.¹⁶ 

Because of the rarity of the cases, standardized treatment for MPN with concurrent BCR-ABL1 fusion and JAK2^V617F mutation has not been established yet. In most clinical scenarios, a combination of kinase inhibitors to target the 2 driver kinases or TKI combined with hydroxyurea has been applied, and it has been adjusted by focusing on the dominant clone when the clonal composition shifts along with the treatment.¹⁵  Although successful anecdotes have been reported, the overall clinical outcome appears to be poor, with approximately half of the patients dying of disease progression.¹⁶  It has been described that MPNs often demonstrate disease progression, such as myelofibrosis, when additional genomic aberration is evident, which may contribute to the poor clinical outcome. Furthermore, comorbidity of composite neoplasms or concurrent genomic alterations complicate the selection of therapeutic targets and prediction of pharmacologic toxicity. In our series, all the patients were treated with a combination of kinase inhibitors, and 3 patients received additional hydroxyurea after the second genomic alteration was identified. One patient (patient 2) died of disease complications soon after the second diagnosis, and 2 other patients (patients 3 and 6) were in very poor condition with unstable hematologic status despite periodic adjustment of their medications, reflecting a difficulty in controlling the disease and a reduced tolerance to pharmacologic side effects. Interestingly, 3 of our patients (patients 1, 4, and 5) eventually received allogenic hematopoietic stem cell transplantation, and all demonstrated a sustained remission with complete donor cell engraftment with undetectable or unquantifiable (patient 1) genomic markers, suggesting an ultimate therapeutic approach at the present time. Hopefully, along with a better understanding of the molecular mechanism of this particular clinicopathologic phenomenon, novel therapeutic targets and treatment strategies may be introduced to provide optimal care for this particular patient population.