THE MOLECULAR DIAGNOSIS OF MYELOPROLIFERATIVE NEOPLASMS
Myeloproliferative neoplasms (MPNs) are a group of conditions that cause abnormal growth of blood cells in the bone marrow. They include polycythemia vera (PV), essential thrombocythemia (ET), primary myelofibrosis (PMF), and other entities such as chronic myeloid leukemia (CML), chronic neutrophilic leukemia (CNL), chronic eosinophilic leukemia not otherwise specified (CEL-NOS) and MPN, unclassifiable (MPN-U). PV, ET, and PMF are further classified as Philadelphia chromosome- or BCR-ABL-negative MPNs.1
The diagnosis of a MPN is suspected based upon clinical, laboratory, and pathological findings (i.e., bone marrow morphology). MPNs are related, but distinct from, myelodysplastic syndromes (MDS). In general, MDS are characterized by morphologic dysplasia and ineffective or dysfunctional blood cells, while MPNs are typically characterized by an increase in the number of blood cells (erythrocytosis, thrombocytosis, and/or leukocytosis).
Classification systems for MPNs include those established by the World Health Organization (WHO) and the International Consensus Classification (ICC).2,3 The ICC was established with input from prior WHO editors and senior advisors, leaders from the Society for Hematopathology (SH) and the European Association for Haematopathology (EAHP).4 Though there is significant overlap between the two classification systems, there are some differences between them. For example, the WHO 5th edition no longer incorporates increased red cell mass as a diagnostic criterion for the diagnosis of PV.2
CLASSICAL BCR-ABL-NEGATIVE MPNs (PV, ET, and PMF)
Polycythemia Vera (PV)
PV is a chronic MPN characterized by increased hemoglobin, hematocrit and red blood cell mass, and with bone marrow biopsy showing panmyelosis.2,3,5,6 Common clinical presentations include pruritis (in nearly half of cases), splenomegaly, and thrombosis; however, patients may also be asymptomatic and the diagnosis may be suspected based on findings from a complete blood count (CBC) performed for other reasons. There is an associated increased risk for thrombosis and transformation to AML or PMF and the most frequent causes of death are leukemic transformation, second malignancies, and thrombotic complications.6
The JAK2 V617F mutation in exon 14 is present in the vast majority of PV, accounting for approximately 95-98% of cases. Functionally similar insertions and deletions in JAK2 exon 12 account for most remaining cases of JAK2 V617F mutation-negative PV. Together, they are identified in 98-100% of PV cases and lead to high diagnostic certainty.6 Absence of a JAK2 mutation, combined with normal or increased serum erythropoietin level, greatly decreases the likelihood of a PV diagnosis.
Though testing for the JAK2 mutations can be serially performed given that the exon 14 mutation is by far the most common, some experts favor targeting both exons 14 and 12 at the same time, in order to avoid undue delays in the diagnostic process; additionally, peripheral blood and bone marrow samples are equally informative in detecting and quantifying JAK2V617F.6 Patients with JAK2 exon 12-mutated PV tend to be younger in age, have increased mean hemoglobin/hematocrit, and lower mean white blood cell and platelet counts at diagnosis compared to those with JAK2 V617F-mutated PV. However, the JAK2 mutations do not confer any differences in prognosis and survival for patients with PV.5
Over 50% of patients with PV harbor DNA sequence variants/mutations other than JAK2, with the most frequent being TET2 (18%) and ASXL1 (15%). Prognostically adverse mutations include SRSF2, IDH2, RUNX1, and U2AF1, with a combined incidence of 5%–10%.6 Though these may be prognostically informative, they are not required for a diagnosis when a JAK2 mutation is identified.
Essential Thrombocythemia (ET)
ET is a disorder of sustained increased platelet count with bone marrow biopsy showing proliferation mainly of the megakaryocytic lineage. Common clinical presentations include hemorrhage and splenomegaly (each occurring in about 30% of patients) though many patients are asymptomatic and, similar to PV, the diagnosis may be suspected based on findings from a complete blood count (CBC) performed for other reasons.7 Hemorrhagic and thrombotic complications occur in <10% to 84% of patients.7
Approximately 50-60% of ET patients carry a somatic JAK2 V617F mutation, while a much smaller percentage (3-5%) have activating mutations in MPL (particularly MPLW515L/K in exon 10). Frameshift mutations in exon 9 of CALR are reported in approximately 20% to 35% of all patients. Another 20% of cases are triple-negative (non-mutated JAK2, MPL, and CALR).5,6
ICC major criteria for the diagnosis of ET include the presence of a JAK2, CALR, or MPL mutation and the exclusion of other MPNs and myeloid neoplasms; minor criteria include the presence of another clonal marker or absence of evidence of reactive thrombocytosis.2,5
Primary Myelofibrosis (PMF)
PMF is characterized by a proliferation of abnormal megakaryocytes and granulocytes in the bone marrow, which is replaced with fibrous tissue, leading to bone marrow failure. However, clinical features are similar to ET and recognizing the prefibrotic stage of PMF is necessary to separate it from ET, PV, and fibrotic PMF.3 The approximate incidence of PMF is 1 in 100,000 individuals. As with the other classic MPNs, patients can be asymptomatic in the early stages of the disease and progression can include transformation to AML. Somatic molecular markers in PMF are similar to those found in ET, and include JAK2 V617F (60%), CALR (~25%), and MPL (5-10%%), among others like trisomy of chromosome 9 and deletion of chromosome 13q.
ICC major criteria for the diagnosis of PMF include presence of a mutation in JAK2, CALR, MPL or presence of another clonal marker (i.e. ASXL1, EZH2, IDH1, IDH2, SF3B1, SRSF2, and TET2 mutations) or absence of reactive bone marrow reticulin fibrosis in the presence of relevant bone marrow findings (i.e. megakaryocytic proliferation and atypia, and appropriate amount of fibrosis for the pre-fibrotic vs fibrotic stages of disease).2,5 As in ET, major criteria also include the exclusion of the other MPNs, MDS and myeloid neoplasms. 2,5 Additionally, molecular testing for AML-associated gene alterations (i.e. mutations in ASXL1, EZH2, TP53, SRSF2, and IDH1 or IDH2; aberrations in chromosomes 1q and 9p) is recommended as part of the initial workup of patients with accelerated/blast phase PMF.5
Transformation and Survival in PV, ET, and PMF
ET transforms into PV in about 2% of patients and both ET and PV can transform into PMF and AML. Overall transformation rates of the three MPNs into leukemia range from about 2 and 4% for ET and PV, respectively, to about 9% for PMF at 20 years.7,8 Patients with MPNs also have a higher risk of other malignancies (i.e. skin, brain, kidney etc.). Median overall survival (OS) in PV, ET, and PMF is about 12 years, 12-14 years, and 4 years, respectively.7,8
Specific driver mutations have been reported to impact rates of leukemic transformation, risk of thrombosis, and overall survival (OS). For example, ET and PMF patients with CALR mutations have a lower risk of thrombosis compared with patients harboring a JAK2 mutation.5,9 However, there is no reported difference in OS or myelofibrotic or leukemic transformation compared to JAK2-mutated ET.5,9 Other mutations are also known to have prognostic impact.
For a list of gene alterations associated with prognostic implications, see Table 1: Prognostic Significance of Mutations in BCR-ABL Negative MPNs.
NON-CLASSICAL BCR-ABL-NEGATIVE MPNs
Chronic neutrophilic leukemia (CNL) is a BCR-ABL-negative MPN characterized by sustained non-reactive peripheral blood neutrophilia, bone marrow hypercellularity due to neutrophilic granulocyte proliferation, and hepatosplenomegaly. Mutations in the CSF3R gene (CSF3R T618I and others) are detected in 60-90% of cases and constitute the diagnostic genetic signature of CNL, though the absence of a CSF3R mutation does not exclude the possibility of CNL.2 However, additional mutations are also seen in most cases, including SETBP1, ASXL1, SRSF2, and other signaling mutations.2 An additional diagnostic criterion for CNL requires the exclusion of BCR-ABL-positive CML, PV, ET, PMF or myeloid/lymphoid neoplasms with eosinophilia and tyrosine kinase gene fusions.2 Most patients with CNL have a poor prognosis, with a mean overall survival of 1.8 years.2
Chronic eosinophilic leukemia, not otherwise specified, (CEL, NOS) is a MPN characterized by persistent eosinophilia that does not meet the diagnostic criteria for any of the other defined hematologic disorders. Bone marrow findings include hypercellularity with dysplastic megakaryocytes and significant fibrosis associated with an eosinophilic infiltrate.2 The diagnosis requires that no tyrosine kinase gene fusions (including BCR-ABL, other ABL1, PDGFRA, PDGFRB, FGFR1, JAK2, or FLT3 fusions) are identified.2
MPN, unclassifiable (MPN-U) is the term used for cases with clinico- morphologic and molecular features of a MPN but that do not satisfy the diagnostic criteria of any of the other defined hematologic disorders or MPN subtypes. This may include patients with a very early-stage untypeable MPN. Such cases must be closely followed such that a specific MPN diagnosis can be rendered as soon as possible.2
SPECIAL CONSIDERATIONS REGARDING TESTING FOR GENE ALTERATIONS IN BCR-ABL NEGATIVE MPN
NCCN guidelines endorse either the single-gene PCR reflex testing approach or the initial multigene approach for the molecular diagnosis of BCR-ABL negative MPNs. For the single-gene approach, guidelines recommend that molecular testing of blood or bone marrow for JAK2 V617F be performed; if negative, it is important to test for CALR and MPL mutations (for patients with suspected ET and PMF) and JAK2 exon 12 mutations (for patients with suspected PV).5 The alternative multigene approach to testing involves the performance of a panel that includes all relevant genes required to establish a diagnosis that can be reasonably assessed by the test. In either case, highly sensitive assays are required for the diagnostic molecular testing of MPNs. Tests detecting JAK2 V617F must have a lower detection limit of <1% VAF and those detecting mutations in CALR and MPL require lower detection limits of 1% to 3%.2,5 Further, in triple-negative cases, it is important to search for noncanonical JAK2 and MPL gene alterations.5
While JAK2, CALR, and MPL mutations are considered driver events in BCR-ABL negative MPN, such driver mutations are not always found, even in the classical MPNs. This is the case for approximately 10% of patients with PMF.5 For such triple-negative cases, NGS may be useful to establish clonality in these or other less commonly associated genes.5 Finding a JAK2 V617F VAF of <1% should also prompt the search for coexisting canonical CALR and MPL mutations.2,5 Further, mutations in other genes – mainly TET2, ASXL1, and DNMT3A – are found in over half of patients with MPN, whereas mutations affecting splicing, cellular signaling and other functions are less common.3 These less common mutations are more frequently observed in PMF and in advanced disease compared to PV and ET and may be prognostically important.3
Once a MPN diagnosis has been confirmed NCCN guidelines recommend testing of additional genes by NGS for mutational prognostication.5 As shown in Table 1 below, alterations in driver and other genes are associated with overall survival (OS), PMF-free survival (for PV and ET), and rates of leukemic transformation.3,5,9 For example, for patients with PMF, survival is longest for CALR(+) ASXL1(-) patients and shortest for CALR(-) ASXL1(+) patients (median survival is approximately 10 vs 2 years, respectively).5 In PV and ET, adverse mutations in SRSF2 (PV) and SRSF2, SF3B1, U2AF1, and TP53 (ET) are included in risk stratification.5
The presence of specific clonal markers can help guide patient management. For example, one of the main goals of management for PV and ET is to identify patients at high risk for thrombosis and prevent complications. For PV and ET, such high-risk individuals include those age >60 years and those with a history of thrombotic event(s). However, as previously discussed, CALR mutations are also used in assessing risk for thrombosis which can be used to inform therapy decisions.5 Similarly, mutations in ASXL1 independently predict inferior survival in PMF and should be included in testing PMF patients planned for stem cell transplant (SCT), the only potentially curative option.9 Mutations in RAS are associated with refractoriness to JAK-inhibitor (JAKi) therapy9 and the presence of high molecular risk mutations or three or more mutations has been associated with inferior response to interferon alfa-2b or peginterferon alfa-2a in patients with PMF.10 In the COMFORT-II study, patients harboring prognostically detrimental mutations (ASXL1, EZH2, SRSF2, IDH1/2) had improved survival and reduced the risk of death when treated with Ruxolitinib.11
In rare circumstances, concurrent alterations in BCR-ABL and JAK2 have been reported. A multi-center study evaluating 1570 patients with suspected MPN found that only 6 (0.4%) tested positive for mutations in both BCR–ABL1 and JAK2 V617F.12 For some patients, the second alteration was identified up to 129 months later; in 5 of these cases, BCR–ABL1 was initially negative for alterations and the mutation in JAK2 was the first to be diagnosed.12 In none of the concurrently mutated cases were there any overt histomorphologic characteristics predictive of the subsequent development of CML.12 The investigators concluded that the co-occurrence of the mutations likely reflects two distinct MPN processes that occur sequentially. Findings that might raise the index of suspicion for the presence of concurrent mutations in BCR–ABL1 and JAK2 include mixed bone marrow cytologic features and unexpected changes in bone marrow histomorphology and laboratory parameters.12
Finally, the clinical utility of quantitating the JAK2 V617F allele burden for the purpose of monitoring disease progression and therapy response is an evolving topic. Allele burden reductions JAK2 V617F have been noted in patients with PMF after ruxolitinib therapy and have also correlated with spleen volume reductions; JAK2 V617F negativity or allele burden reduction after allogeneic HCT has also been associated with a decreased incidence of relapse.13-16 Additionally, a high JAK2 allele burden has been reported as a risk factor for myelofibrotic transformation and thrombotic events in patients with PV and ET. A multicenter study of 347 JAK2V617F-positive patients found that those with a persistently high (≥50%) or unsteady JAK2V617F load at follow-up had an increased risk of myelofibrotic transformation and a trend for a higher incidence of thrombosis (incidence rate ratios [IRR] 20.7 and 1.7, respectively) than patients with a stable allele burden below 50%.17 Therefore, monitoring of JAK2V617F allele burden may be useful in patients with MPN for predicting disease's complications, particularly myelofibrotic transformation.17 It may also be useful to help assess the impact of molecular response to therapy. However, the utility of JAK2 V617F allele burden reduction as a predictor of clinical outcome is not well-established and NCCN guidelines do not currently recommend the routine monitoring of JAK2 V617F allele frequency to guide treatment decisions.5 The International Working Group-Myeloproliferative Neoplasms Research and Treatment (IWG-MRT) and European LeukemiaNet (ELN) consensus report also do not require cytogenetic and molecular responses for assignment of complete response in PMF.18
Table 1: Prognostic Significance of Mutations in BCR-ABL Negative MPNs (adapted from NCCN Myeloproliferative neoplasms, version 2.2023 Tables MPN-E5 and Vannucchi and Guglielmelli 20229)
Mutated Gene |
Polycythemia Vera |
Essential Thrombocythemia |
Primary Myelofibrosis |
JAK2 |
JAK2 V617F and JAK2 exon 12 mutation: Similar rates of thrombosis, evolution to PMF or leukemia, and death. |
JAK2 V617F: Higher risk of thrombosis |
JAK2 V617F: Intermediate prognosis and higher risk of thrombosis than patients with mutations in CALR. |
CALR |
|
Lower risk of thrombosis compared to JAK2-mutated ET.
No difference in OS or myelofibrotic or leukemic transformation compared to JAK2-mutated ET.
|
Lower risk of thrombosis compared to JAK2 mutation.
Improved OS compared to JAK2 and "triple-negative" PMF.
|
CALR Type 1/Type 1-like |
|
|
Improved OS compared to CALR type 2/ type 2-like and JAK2 V617F mutation. |
MPL W515L/K |
|
|
Lower hemoglobin levels at diagnosis and increased risk of transfusion dependence.
Intermediate prognosis and higher risk of thrombosis compared to patients with CALR mutation. |
"Triple Negative" (non-mutated JAK2, MPL, and CALR) |
|
|
Inferior LFS compared to patients with JAK2 and/or CALR-mutated PMF.
Inferior OS compared to patients with CALR-mutated PMF. |
TP53 |
|
Inferior LFS |
Associated with leukemic transformation. |
ASXL1, IDH1/2, RUNX1, SRSF2 |
The presence of at least 1 is associated with inferior OS. ASXL1: Inferior PMF-free survival; IDH2 and RUNX1: Inferior LFS |
|
|
ASXL1, EZH2, IDH 1 and 2, RAS, SRSF2, U2AF1 Q157 |
|
|
ASXL1, EZH2 RAS, SRSF2, U2AF1 Q157: Inferior OS. ASXL1, IDH 1 and 2, SRSF2: Inferior LFS; ASXL1, IDH 1 and 2: Inferior PFS following HCT. |
U2AF1 or DNMT3A or CBL |
|
|
U2AF1 or DNMT3A or CBL: worse OS in patients undergoing allogeneic HCT. |
EZH2, IDH2, RUNX1, SH2B3, SRSF2, SF3B1, TP53, U2AF1
|
|
The presence of at least 1 is associated with inferior OS. U2AF1 and SF3B1: Inferior PMF-free survival; EZH2 and RUNX1: Inferior LFS |
|
NOTE: Where there is no information provided in the table, NCCN guidelines have not attributed prognostic significance for a particular gene mutation for the given disorder. LFS: leukemia-free survival; PFS: progression-free survival, OS: overall survival