Prognosis of Primary Myelofibrosis in the Genomic Era

Currently, prognostication in primary myelofibrosis (PMF) relies on the International Prognostic Scoring System (IPSS), dynamic IPSS (DIPSS), and DIPSS-plus, which incorporate age, blood counts, constitutional symptoms, circulating blasts, red cell transfusion need, and karyotype. Although the JAK2 V617F mutation was discovered a decade ago and MPL mutations shortly thereafter, it was the recent discovery of CALR mutations in the vast majority of JAK2/MPL-unmutated patients and recognition of the powerful impact of CALR mutations and triple-negative (JAK2/MPL/CALR-negative) status on outcome that set the stage for revision of traditional prognostic models to include molecular information. Additionally, the advent of next-generation sequencing has identified a host of previously unrecognized somatic mutations across hematologic malignancies. As in the myelodysplastic syndromes, the majority of common and prognostically informative mutations in PMF affect epigenetic regulation and mRNA splicing. Thus, a need has arisen to incorporate mutational information on genes such as ASXL1 and SRSF2 into risk stratification systems. Mutations in yet other genes appear to be important players in leukemic transformation, and new insights into disease pathogenesis are emerging. Finally, the number of prognostically detrimental mutations may affect both survival and response to ruxolitinib, which has significant implications for clinical decision making. In this review, we briefly summarize the prognostic models in use today and discuss in detail the somatic mutations commonly encountered in patients with PMF, along with their prognostic implications and role in leukemic transformation. Emerging prognostic models that incorporate new molecular information into existing systems or exclude clinical variables are also presented.