Secondary Mutations of JAK3 and SETBP1 in Juvenile Myelomonocytic Leukemia and Their Propagating Capacity; A Report from the AIEOP Study Group

INTRODUCTION

Juvenile myelomonocytic leukemia is a rare early childhood leukemia, characterized by excessive proliferation of granulocytic and monocytic cells. About 95% of JMML patients harbor driver mutations in the RAS signaling pathway. Recently, secondary hits in SETBP1 and JAK3 have been reported in a Japanese cohort of JMML patients showing an adverse clinical outcome of patients carrying these mutations. Here we report the mutational analysis of SETBP1 and JAK3 and clinical implications in a cohort of Italian JMML patients.

METHODS

Samples collected at diagnosis of 65 patients with JMML were analyzed by Sanger sequencing. Mutations were found in RAS (NRAS-KRAS) 31%, PTPN11 35%, CBL 5%, whereas in 29% of patients none of the above cited mutations was present. Mutation hot spot regions of SETBP1 (SKI domain) and of JAK3 (PTK domains) were sequenced. A xenografted murine model was used to assess the in vivo competitive repopulation advantage of clones carrying mutations of JAK3 and SETBP1. Mononuclear cells from a patient with JMML at diagnosis harboring PTPN11, SETBP1 and JAK3 mutations were transplanted in NSG mice and assessed for mutational status in the bone marrow and spleen after engraftment of JMML cells.

RESULTS

Screening for JAK3 and SETBP1 mutations in patients revealed 9 mutations in 8 out of 65 patients at diagnosis of JMML. All of the identified secondary mutations were associated with known driver mutations, more frequent with mutated PTPN11 and RAS (p=0.036 and p= 0.01 respectively) than with CBL or in cases without known driver mutations. Seventy-five percent of secondary mutations were found in SETBP1 and only 1 patient harbored a mutation in JAK3. Remarkably one patient carried mutations in JAK3 (L857P and L857Q, both predicted to damaging protein function), PTPN11 (G503A) and SETBP1 (D868N). All variants were identified as heterozygous mutations, confirmed bi-allelic expression at the transcriptome level. The only patient carrying JAK3 as secondary mutation at E958K showed wild-type expression of JAK3 pointing to absence of a functional role at the protein level. Univariate analysis revealed association between the presence of secondary mutations and patient’s age at diagnosis, with older patients carrying JAK3 and SETBP1 mutations (p=0.0067); no other clinical and biological characteristics (i.e. WBC count, percentage of monocyte, HbF level and platelet count) being significantly associated with the presence of secondary hits in bone marrow of JMML cases. Patients with secondary mutations showed a trend to shorter survival compared to those without secondary events in JAK3 and SETBP1 (5-years OS= 0% vs 54.01%, SE=8.1; p=0.41, respectively). Interestingly, the in vivo assay using xenografted mice revealed a different propagating capacity of JAK3 clones of patients carrying JAK3 (2 different clones), SETBP1 and PTPN11 mutations. Indeed, for JAK3 only the clone with the L857Q mutation engrafted in BM and spleen of the mouse, together with SETBP1 and PTPN11 mutations. Moreover, a second mouse engrafted with mononuclear cells of the same patients showed that only cells carrying the PTPN11 mutation had engrafted.

CONCLUSIONS

In conclusion we identified secondary mutations in JAK3 and SETBP1 in 12% of patients of a representative cohort of Italian JMML patients, showing a trend of adverse outcome for patients carrying these mutations. These secondary events in JMML patients showed to have distinct propagating capacities upon engraftment in NSG mice pointing to a different functional impact of these mutations.