Life, genes, and death in Ph⁻ MPNs

In this issue of Blood, Tefferi et al make the important observations that polycythemia vera (PV) is not a continuum from essential thrombocythemia (ET), that survival in ET is less than matched controls but of longer duration than in patients with PV and primary myelofibrosis (PMF), and that “triple negative” mutational status in PMF is an important adverse risk factor for blast transformation. Genetic profiling should be integrated into classical methods for profiling these diseases.¹ 

Even before William Dameshek described the interrelationship of the myeloproliferative diseases, hematologists noted their frequent termination in a phenotypic picture resembling acute myeloid leukemia.²  The discovery of the Philadelphia chromosome (Ph⁺) permitted the division of this group of diseases into Ph⁺ chronic myeloid leukemia (CML) and the main Ph⁻ subgroup encompassing PV, ET, and PMF. The development of new chemotherapeutic agents and their enthusiastic use beginning in the 1950s for all types of leukemia, but especially acute leukemia, required a reliable demarcation of the blast phase of CML. Nearly a half century ago, this was based on the presence of the Ph⁺ chromosome and the percent blasts in blood or marrow,³  a set of criteria which has survived the test of time, and is applicable to the Philadelphia chromosome negative myeloproliferative neoplasms (Ph⁻ MPNs) as well. Little interest existed in identifying other risk factors because the chemotherapeutic results of blast-phase Ph⁻ MPNs were even less satisfactory than those of patients with CML, especially after the advent of the tyrosine kinase inhibitors.

The discovery of the Janus kinase 2 (JAK2) mutations⁴  (seen in nearly all patients with PV and in slightly more than half of the patients with ET and PMF) and the myeloproliferative leukemia virus oncogene (MPL) mutation⁵  (in 5%-10% of patients with ET and PMF) rekindled enormous interest in the Ph⁻ MPN diseases. Clinical and basic science studies led to an improved understanding of these disorders. It was suggested that ET and PV were a continuum of one basic disease.⁶  Studies of risk factors for treatment of the Ph⁻ MPNs assumed importance, as frenzied activity by pharmaceutical companies followed in order to develop inhibitors of JAK2 mutations. The most recently discovered genetic abnormality, calreticulin (CALR) exon 9 insertions/deletions,⁷  which presumably upregulates JAK2 activity, has been found in about 25% of patients with ET and in 30% of patients with PMF, respectively, but not in PV. Thus, CALR affords another way to clinically delineate one Ph⁻ MPN from another.

In a most timely and important article, Tefferi et al have assessed in perspective the clinical and biological importance of JAK2/MPL/CALR mutational studies. They have combined nearly 1600 patients from the Mayo Clinic (Rochester, MN) and centers in Florence, Italy, and Bergamo, Italy, analyzing their phenotypic and genotypic characteristics especially regarding survival and blast transformation. They offer evidence that ET and PV are not a continuum of one disease, but are separate entities based upon morphologic and clinical analysis. Survival of ET patients is prolonged but not equal to that of a matched normal population, but longer than that of patients with PV and PMF. Of the various permutations of the 3 genes, “triple-negative” PMF patients seem to be a breed apart and bear special watching for the development of blast crisis. The authors wisely advise that their analyses should be interpreted and interdigitated with other observations such as the Dynamic International Prognostic Scoring System plus⁸  and other identified mutations associated with an adverse prognosis such as the ASXL1 gene.

Some caveats exist. Despite the large number of patients, the results should be considered “preliminary” because the study was retrospective, based upon 2008 World Health Organization (WHO) criteria⁹  for which some reservations have been expressed, especially in the case of PV.¹⁰  Patient selection for mutation analysis rested upon availability of archival DNA. Duration of follow-up differed between the Mayo and Italian patients as did the number of patients followed to death.

Nevertheless, Tefferi et al make the relevant points that the diseases remain distinguished not only by their genetic profile but also by their phenotypic characteristics and clinical course. Moreover, mutations have not supplanted the need for morphologic and otherwise classical methods for disease distinction which have prognostic and therapeutic relevance. Thus, the final word has not been written but this article is important because it emphasizes the need for correlating the new genetic markers with the classical methods, including examination of marrow biopsies, for evaluating these diseases.