Incidence and Impact of Activating Mutations of the RAS-RAF-MEK-ERK Pathway in Patients with Philadelphia-Negative (Ph-negative) Myeloproliferative Neoplasms (MPNs) and Myelodysplastic/Myeloproliferative Disorders–Unclassified (MDS/MPD-U)

Introduction: Mutations that activate the RAS-RAF-MEK-ERK pathway have long been known to occur in patients with solid tumors and hematological malignancies. The most common mutations occur in the Ras family of GTPases (HRAS, NRAS, KRAS) and the Raf family of serine-threonine kinases (ARAF, BRAF, CRAF). In myeloid malignancies, RAS mutations have mainly been described in patients with acute myeloid leukemia, chronic myelomonocytic leukemia (CMML) and myelodysplastic syndrome. There are few studies describing the incidence of mutations of the RAS-RAF-MEK-ERK pathway in patients with MPNs other than CMML.

Objective: To describe the incidence, clinical features and prognostic impact of Ras and Raf mutations in patients with Ph-negative MPNs and MPN/MDS-U

Methods: Paired DNA (sorted CD66b-granulocytes/skin biopsy) from patients with MPNs or MPN/MDS was subjected to whole exome sequencing on a Illumina HiSeq 2000 platform using Agilent SureSelect kit (see our abstract “Whole Exome Sequencing of Myeloproliferative Neoplasms and Myelodysplastic/Myeloproliferative Disorders”). Tumor coverage was 150x and germline coverage was 60x. Somatic variants calls were generated by combining the output of Somatic Sniper (Washington University), Mutect (Broad Institute) and Pindel (Washington University), followed by in-house filters to reduce false positive calls. Statistical calculations were done in Stata, v11.0.

Results: We found clonal activating mutations of the RAS-RAF-MEK-ERK pathway in 8 patients (6.7% of cases). Diagnosis included primary myelofibrosis (PMF; N=5), MDS/MPD-U (N=2) and essential thrombocythemia (ET; N=1). Their clinical features are summarized in Table 1 (three of these patients [UPIs #11, #13, #99] are also described in the abstract “Genomic Profile of Patients with Triple Negative (JAK2, CALR and MPL) Essential Thrombocythemia and Primary Myelofibrosis”). There were 7 NRAS mutations and 1 BRAF mutation. In 5 cases the variant allele fraction (VAF) of reads in the tumor sample indicated that the mutation was present in a subclone at the time of sequencing. We next compared the clinical features of these 8 patients with 79 patients (MF=43, ET=35, MDS/MPD=1) who did not harbor these mutations. Patients with NRAS/BRAF mutations had lower hemoglobin (8.3 vs. 11.8 g/dL, p=0.001), higher white blood cell counts (28.37 vs. 7.7 x10⁹/L, p=0.008) and had higher lactate dehydrogenase (1041 vs. 685 IU/L, p=0.02). They also had worse overall survival compared to unmutated cases (Hazard ratio [HR]=11.57; p=0.001). Most patients with NRAS/BRAF mutations had a high number of concomitant driver mutatons (median 5 vs. 1; p<0.0001). When the number of driver mutations was analyzed together with NRAS/BRAF mutations in a Cox model, NRAS/BRAF mutations were no longer independent predictors of survival (HR=1.48; p=0.61).

Conclusions: Activating mutations of the RAS-RAF-MEK-ERK pathway occur in 6-7% of patients with Ph-negative MPNs, and they tend to co-occur with a high number of concomitant driver mutations. In most cases the mutation was present in a subclone, suggesting that they are late occurring. Patients with NRAS/BRAF mutations had a trend for worse outcome, but that was mainly dependent on the total number of driver mutations. The activity of MEK and BRAF inhibitors needs to be explored in patients with Ph-negative MPNs who harbor activating mutations of the RAS-RAF-MEK-ERK pathway.