Sequential Analysis By Next Generation Sequencing of 18 Genes in Polycythemia Vera and Essential Thrombocythemia Reveals a Association Between Mutational Status and Clinical Outcome after 3-Year Follow-up

Introduction

Myeloproliferative neoplasms (MPN) are molecularly characterized by driver mutations of JAK2, MPL or CALR. Other somatic mutations may occur in epigenetic modifiers or oncogenes. Some of them have been shown to confer a poor prognosis in primary myelofibrosis, but their impact is less known in Polycythemia Vera (PV) and Essential Thrombocythemia (ET). In this study, we investigated the mutational profile using NGS technology in 50 JAK2V617F positive cases of MPN (27 PV and 23 ET) collected at the time of diagnosis and after a 3 year follow-up (3y).

Patients and Methods

All patients were JAK2V617F positive and already included in the prospective cohort JAKSUIVI. All exons of JAK2, MPL, LNK, CBL, NRAS, NF1, TET2, ASXL1, IDH1 and 2, DNMT3A, SUZ12, EZH2, SF3B1, SRSF2, TP53, IKZF1 and SETBP1 were covered by an Ampliseq^TM custom design and sequenced on a PGM instrument (Life Technologies). CALR exon 9 mutations were screened using fragment analysis. Hotspots that mutated recurrently in MPN with no sequencing NGS coverage were screened by Sanger sequencing and HRM. A somatic validation was performed for some mutations using DNA derived from the nails. The increase of a mutation between diagnosis and follow-up has been defined as a relative increase of twenty percent of the allele burden. An aggravation of the disease at 3y was defined by the presence of at least one of the following criteria: leukocytosis >12G / L or immature granulocytes >2% or erythroblasts >1%; anemia or thrombocytopenia not related to treatment toxicity; development or progressive splenomegaly; thrombocytosis on cytoreductive therapy; inadequate control of the patient’s condition using the treatment (defined by at least one treatment change for reasons other than an adverse event).

Results

As expected, the JAK2V617F mutation was found in all patients with the use of NGS. In addition, we found 27 other mutations in 10 genes out of the 18 genes studied by NGS (mean 0.54 mutations per patient). Overall, 29 of 50 patients had only the JAK2V617F mutation and no other mutation in any of the genes analysed. No CALR mutation was detected. Nine mutations that were not previously described in myeloid malignancies were found. The genes involved in the epigenetic regulation were those most frequently mutated: TET2, ASXL1, IDH1, IDH2 and DNMT3A. In particular, TET2 mutations were the most frequent and occurred in 20% of cases. There was no difference in the number or in the presence of mutations between PV and ET.

At 3y, 4 mutations appeared in 4 patients and 15 out of 50 patients (9 PV and 6 ET) were affected by an allele burden increase of at least one mutation. At 3y, 24/50 patients suffered an aggravation of the disease as defined by the primary outcome criterion (16 PV and 8 ET). The presence of a mutation (JAK2V617F omitted) at the time of the diagnosis was significantly associated with the aggravation of the disease (p=0.025). Retaining only mutations with an allele burden greater than 20%, the association with disease aggravation is more significant (p=0.011). Moreover, a mutation of ASXL1, IDH1/2 or SRSF2, which is a poor prognostic factor in primary myelofibrosis, was found in 8 patients, all having presented an aggravation of their disease (p=0.001). Only 4 patients had more than one somatic mutation other than JAK2V617F and all of them also had an aggravation at 3y (p=0.046). In this cohort, appearance of a mutation at 3y was not associated with the course of the disease. Conversely, the increase of allele burden of at least one mutation was associated with an aggravation (p=0.019).

Discussion and conclusion

Despite the short follow-up and the limited number of patients, this study suggests that the presence of additional mutations at the time of the diagnosis in PV and TE is correlated to a poorer disease evolution. The increase of mutation allele burden, which reflects clonal evolution, also seems to be associated with the course of the disease. These results argue for a clinical interest in large mutation screening by NGS at the time of the diagnosis and during follow-up in ET and PV.