NEW YORK (GenomeWeb) – Researchers have uncovered distinct genetic subtypes of myeloproliferative neoplasms, which, in combination with other clinical data, could help guide treatment approaches.
Myeloproliferative neoplasms are a group of chronic blood cancers associated with a risk of bleeding and blood clots, but they can also progress to more advanced disease, including leukemia. The current classification system divides these blood cancers into groups based on their clinical and laboratory features, but there is some overlap between the categories.
In order to classify the cancers molecularly, a team led by researchers at the Wellcome Trust Sanger Institute examined driver mutations and copy number changes in more than 2,000 patients with myeloproliferative neoplasms. As they reported in the New England Journal of Medicine yesterday, they uncovered eight distinct genetic subtypes with distinct clinical profiles and bundled those with other variables to develop a prognostic model.
"Our new online calculator takes genetic and clinical information available for a patient and makes a prediction of the future outcome of that particular person's disease," Jyoti Nangalia, a researcher at the Wellcome Sanger Institute and one of the lead authors, said in a statement. "In the future, this could be used to reassure patients who have a good predicted outcome, and identify patients who are at risk of developing severe disease who could benefit from an alternative treatment approach."
The researchers performed targeted sequencing — focusing on 69 genes previously linked to myeloid cancers — and genome-wide copy number analysis on 1,887 patients, and studied data from 148 patients who previously underwent whole-exome sequencing. These patients included individuals with essential thrombocythemia, polycythemia vera, myelofibrosis, and other myeloproliferative neoplasms.
Thirty-three genes harbored driver mutations in at least five patients. JAK2, MPL, and CALR accounted for 1,831 driver mutations and were, for 45 percent of patients, the only driver mutation present.
More than 1,000 driver mutations affected other genes. For instance, the researchers uncovered 45 truncating mutations in PPM1D in 38 patients and nonsense or frameshift mutations affecting MLL3 in 20 patients.
Some mutations, the researchers noted, tended to arise in patients with particular disease subtypes. The JAK 46/1 haplotype, for instance, was correlated with polycythemia vera, while mutations affecting spliceosome components, epigenetic regulators, and the RAS pathway were linked to accelerated phase disease like myelofibrosis.
Using Bayesian modeling, the researchers identified genetic alterations that could distinguish groups of myeloproliferative neoplasms. In all, they teased out eight subgroups, one of which was marked by TP53 mutations. These mutations often appeared along with changes at chromosome 17p and deletions at chromosome 5q, and were linked to an increased risk of transformation to acute myeloid leukemia. Meanwhile, another subgroup was defined by mutations affecting chromatin and spliceosome regulators, among other changes, and was associated with increased risk of transformation to myelofibrosis.
The researchers noted that clinical and lab factors were also associated with disease prognosis, and combined 63 clinical and genomic variables into a multivariate statistical model. They cross-validated the model using both an internal and an external cohort to uncover a concordance of 76 percent to 86 percent for overall survival, event-free survival, and transformation to acute leukemia.
This model, the researchers reported, performed better than existing ones. In particular, they said, it was able to bring greater granularity to intermediate-risk patients. It also identified a portion of patients with chronic-phase myeloproliferative neoplasms who were at risk of disease progression. Patients like these, the researchers said, could be considered for clinical trials of new therapeutic agents.
At the same time, they were able to identify a subgroup that was at low risk of disease progression, who might not benefit from aggressive treatment and instead could be treated more conservatively.