The Hematologist

November-December 2017, Volume 14, Issue 6

From Blurred Lines to Guidelines for Myeloproliferative Neoplasms

Aaron T. Gerds, MD, MS Assistant Professor of Medicine, Hematology and Medical Oncology
Cleveland Clinic Taussig Cancer Institute, Cleveland, OH
Ruben Mesa, MD, FACP Director
UT Health San Antonio Cancer Center, San Antonio, TX

Published on: October 27, 2017

When reflecting on guideline development and implementation it is difficult not to lean on the cliché crutch of Yogi Berra’s simple sage advice about knowing where you are going and whether or not you will get there. With a prevalence in the United States of approximately 300,000, myeloproliferative neoplasms (MPNs) are rare, and the lines between them are blurred, with significant overlap in the presentation and clinical management of polycythemia vera (PV), essential thromocythemia (ET), and myelofibrosis (MF). Having a guide to help navigate the subtleties of treatment can be helpful to those who may see only a few cases each year. Another driver in the development of consensus guidelines is the rapid shift from volume-based to valuebased care. Guidelines and carepaths are already being used by hospital systems to increase the value of their care, and insurers have long used guidelines in instances of prior authorizations and reimbursement.

In addition to individual review articles, several non-U.S. groups have published consensus guidelines, including the British Committee for Standards in Haematology, International Working Group for MPN Research and Treatment (IWG-MRT), and European LeukemiaNET (ELN). Like politics, however, patient care is local, and there was a need for U.S.-based guidelines.1 This need was answered in 2016 when the inaugural MPN guidelines were published by the National Comprehensive Cancer Network (NCCN).2 While the first edition covered the diagnosis and treatment of MF, the newest iteration published in September 2017 includes algorithms and a discussion for PV and ET.

The MPN guidelines are in a format that will be familiar to those who have used NCCN guidelines before, and provide updates that may not be familiar to all. One such example is the risk stratification for ET. As trainees, many of us were brought up on the classic thrombohemorrhagic model for risk stratification, in which those patients younger than 60 years who have never had a thrombosis are considered low-risk, and all others are high-risk. Low-risk patients were relegated to daily aspirin alone, while the high-risk were given cytoreductive therapy. However, we know that there are other factors that influence the risk of thrombosis in these patients, such as JAK2  mutational status. Therefore, the panel elected to build its ET management algorithm using the revised International Prognosis Score in Essential Thrombocythemia-Thrombosis (IPSETThrombosis) model (Figure).

The panel also took to task the prognostic relevance of other mutations in PV and ET (Table). Next generation sequencing (NGS) panels are becoming cheaper, faster, and more readily available. Patients, through robust social networks and information sharing, are aware of these tests, and are requesting them. Although, the prognostic significance of many mutations is unknown, and actionable targets are limited, many practicing hematologists are now faced with interpreting these panels on a day-to-day basis.

Although consensus opinion still plays a role in the formation of guidelines, quality research in the field continues, and evidence-based practices are becoming more commonplace. With that in mind, the panel members evaluated the level of evidence on which each recommendation was based. This is done with respect to the efficacy, safety, and affordability of the regimen as well as the quality and consistency of the published evidence. This is then translated into a visual tool called Evidence Blocks™ and published with the guideline. Fortunately, many ongoing trials have the potential to change practice, such as the MPD-RC 112 study, which randomly assigned high-risk PV and ET patients between pegylated interferon-α-2 and hydroxyurea, and for which the results will need to be rapidly assessed and incorporated into the guidelines.

Polycythemia Vera (PV)
ASXL1/SRSF2/IDH1/2The presence of at least one of these "adverse variants/mutations" is associated with inferior overall survival (compared to other sequence variants/mutations, or none), which was independent of age, International Working Group (IWG) prognostic model for PV, and karyotype.

Adverse variants/mutations also affect myelofibrosis-free survival.
JAK2 exon 12 mutationPatients with JAK2 exon 12-mutated PV are of younger age, have increased mean hemoglobin/hematocrit, and have lower mean white blood cell and platelet counts at diagnosis compared to those with JAK2 V617F-mutated PV. However, both JAK2 mutations are associated with similar rates of thrombosis, evolution to myelofibrosis or leukemia, and death.
Essential Thrombocytopenia (ET)
CALRLower-risk of thrombosis compared to JAK2-mutated ET

No difference in overall survival or myelofibrotic or leukemic transformation compared to JAK2-mutated ET

CALR mutation does not modify the IPSET score for predicting thrombosis in patients with ET.
TP53Associated with inferior leukemia-free survival in multivariate analysis
SH2B3/IDH2/U2AF1/SF3B1/EZH2/TP53The presence of at least one of these "adverse variants/mutations" is associated with inferior overall survival (compared to other sequence variants/mutations, or none) independent of age and karyotype.

Adverse variants/mutations also affect myelofibrosis-free survival.

While the lines between the MPNs may be blurred, evidence-based consensus guidelines can help bring things into focus. The expansion of the U.S.-based NCCN guidelines to include the diagnosis and treatment of PV and ET is a big step forward. Of course, they are far from perfect, so the question is: How can we improve them?

Please excuse the back-of-the-envelope exercise here, but a simple search using the term “polycythemia vera” yields 230 hits, whereas a search for “breast cancer” yields 7,701 hits (accessed from on 19 Sep 2017). If you adjust this for the U.S. prevalence of PV (148,000) and breast cancer (3,327,000), there are approximately 155 studies per 100,000 US patients with PV, and approximately 231 studies per 1,000,000 U.S. patients with breast cancer. Although neither overly scientific nor rigorous, this method does highlight the need for more studies. Evidence generated from clinical and translational studies serves as the building material for future additions to the guidelines. Of course we are referring to quality evidence with thoughtful inclusion criteria and meaningful endpoints, but that is a discussion for another time.

Author's Note: These guidelines meet a majority of the Institute of Medicine (now National Academy of Medicine) standards for guideline development in the domains of transparency, management of conflicts, composition of panels, evidence base, articulation of recommendation, external review, and updating. The NCCN Guidelines are not based on systematic reviews because of the numbers of recommendations in any single Guideline, and the goal of continuous updating to reflect evolving data. In particular, they are continuum of care guidelines; therefore, it is not possible to perform systematic reviews on each of the many recommendations contained in a single set of guidelines and maintain their currency. Moreover, the panel regularly reviews relevant published therapeutic trials and peer-reviewed abstracts from scientific meetings, updating the guidelines up to nine times per year. Systematic reviews would hinder this frequency of updates. Furthermore, there are an increasing number of partnerships between industry and academics, particularly in rarer diseases like MPNs. Panel members and staff are required to disclose potential conflicts of interest, which are available online, and are excused from deliberations and voting if warranted by the level of conflict. NCCN panel chairs and vice chairs are permitted to participate in funded scientific programs (research and scientific advisory boards), but not in any type of promotional activity.


  1. Stein BL, O'Brien S, Greenberg P, et al. The need for United States-based guidelines for myeloproliferative neoplasms. J Natl Compr Canc Netw. 2015;13:607-609.
  2. Mesa R, Jamieson C, Bhatia R, et al. Myeloproliferative Neoplasms, Version 2.2017, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2016;14:1572-1611.

Conflict of Interests

Dr. Gerds and Dr. Mesa indicated no relevant conflicts of interest. back to top