Prognostic models in the myeloproliferative neoplasms

The management of myelofibrosis (MF) is predominantly supportive, with the use of JAK2 inhibitors or allogeneic stem cell transplantation reserved for patients predicted to have poor overall survival. Identification of these patients is aided by a number of prognostic scoring systems, foremost among them the Dynamic International Prognostic Scoring System (DIPSS). Similarly, the use of cytoreductive therapies in essential thrombocytosis (ET) and polycythemia vera (PV) is targeted to patients identified as at highest risk of thrombosis. In this context, age and history of prior thrombosis are the strongest risk factors. Several studies have sought to improve the accuracy of prognostic modelling by encorporating a wider range of clinical and genomic variables, while also assessing those predictive for other endpoints, such as disease transformation. This review aims to summarise and evaluate existing prognostic models in the myeloproliferative neoplasms, and examine ways in which they can be further refined.