In classical Philadelphia-negative myeloproliferative neoplasms (MPN), JAK2V617F is considered as a driver mutation when the threshold of 1% JAK2V617F/JAK2total allele burden is reached. However a lower ratio is sometimes detected with highly sensitive assays. We investigated the clinical significance of such minor clones by describing the characteristics of 27 patients with a low JAK2V617F allele burden ranging from 0.1% to 0.99%.
A commercially available quantitative ASO-PCR assay of 0.1% sensitivity (MutaQuant® kit, Ipsogen) was used. Two thousand five hundred consecutive blood samples were sent to our lab for JAK2V617F mutation between 2009 and 2012. Total blood DNA was extracted by an automated standardized procedure (Qiasymphony®, Qiagen). All samples were tested in duplicate. The 27 samples of our cohort were controlled using a second assay of 0.01% sensitivity (Larsen et al, BJH 2007). Thirty samples from healthy donors were also tested. High resolution melting curve (HRM) analysis of JAK2 exon 14 ruled out the possibility of an additional mutation hampering the annealing of a primer. Patients with a known classical MPN clinical phenotype were also tested for JAK2 exons 12–17 (entire pseudo-kinase domain) or for MPL exon 10 depending on the context.
Among the 2500 samples, 735 (29.4%) were positive above 1%, 27 (1.1%) had low JAK2V617F allele burden ranging from 0.12 to 0.99%. The patient with the lowest ratio (0.12%) was not confirmed by the second assay and therefore was excluded from the study. This allowed the median to settle at 0.40%. No associated mutations were found in the JAK2 pseudo-kinase domain in patients with polycythemia vera (PV) and in MPL exon 10 in patients with essential thrombocytosis (ET) and primary myelofibrosis (PMF). Healthy patients were all tested JAK2V617F negative.
The cohort included 19 men and 7 women ranging from 28 to 95 years of age (median 63 years old).
Two patients had secondary acute myeloid leukaemia following JAK2V617F positive MPN indicating the presence of residual JAK2V617F cells and the negativity of the myeloblastic population.
Thirteen patients (50%) had a classical MPN with a median ratio of 0.36%: 7 ET, 5 PV and 1 PMF according to WHO 2008 criteria. However a bone marrow biopsy was available for only two patients (1 ET, 1 PMF). None of them had received pegylated interferon alpha-2a. Four patients had a prior history of thrombosis: two strokes, one pulmonary embolism, two portal vein thrombosis (PVT). For one PV patient, a 6 months follow-up blood and bone marrow sample confirmed a low allele burden in the same range (0.4%) and in vitro Epo-independant erythroid colonies were observed.
Five patients had other chronic myeloid neoplasms (two myelodysplastic/myeloproliferative neoplasms, one chronic eosinophilic leukaemia, one chronic myeloid leukaemia, one refractory anaemia with ring sideroblasts). Among these five, four had an abnormal karyotype. We did not observe any thrombotic event in these patients.
We cannot conclude on hematological diagnosis for the last six patients: four patients were screened for JAK2 mutation because of PVT. One patient had chronic polycythemia in a context of alcohol and tobacco abuse. One patient had homozygous hemochromatosis with a normal haemoglobin level in spite of repeated phlebotomies.
In this single centre study low JAK2V617F allele burden represented 1% of all samples sent for JAK2V617F study and 3.5% of JAK2V617F positive patients.
Seventeen patients (65%) had classical MPN or splanchnic vein thrombosis. To our knowledge PV patients with such low JAK2V617F allele burden have not been reported in the absence of associated JAK2 pseudo-kinase domain mutation. A larger screen for cooperating mutations responsible for the PV phenotype is under process.
In the context of other chronic myeloid neoplasms, the JAK2V617F mutation is thought to belong to a more complex clonal architecture mostly implicating chromatin remodeling genes. Here, the presence of a JAK2 mutation could argue in favour of clonal haematopoiesis.
In conclusion the clinical phenotype of low JAK2V617F patients overlaps with classical JAK2V617F MPN. The technical implications might be challenging for molecular diagnostic platforms. More data are needed to further characterize these patients.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.