Next-Generation Sequencing Shows Marked Potential for Diagnostic Improvements in MPNs

Article

Patients with a suspected or diagnosed myeloproliferative neoplasm should undergo targeted next-generation sequencing versus sequential or allele specific testing to reduce the risk of identifying a false driver mutation or an excluded diagnosis of MPN.

Ella Thompson, PhD

Patients with a suspected or diagnosed myeloproliferative neoplasm (MPN) should undergo targeted next-generation sequencing (NGS) versus sequential or allele specific testing to reduce the risk of identifying a false driver mutation or an excluded diagnosis of MPN, according to a study presented during the virtual 2020 European Hematology Association Congress.

In the analysis, the variant allele frequency ratio for CALR/MPL co-mutations of JAK2 V617F was significantly higher, and non-canonical JAK2 and MPL mutations were identified.

“We have identified multiple subgroups of driver mutational profiles in a large cohort of patients with MPN that risk incorrect diagnostic classification if evaluated using sequential or allele specific testing strategies,” the study authors wrote in their poster presentation. “These subgroups are efficiently identified using targeted NGS, and we recommend that this approach should be standard of care in the investigation of MPNs.”

The presence of JAK2, CALR, or MPL mutations are important diagnostic criteria for diagnoses of polycythemia vera, primary myelofibrosis, and essential thrombocythemia (ET). Testing for these genetic drivers is generally performed sequentially, such that if JAK2 Val617Phe is negative, then JAK2 exon 12 or CALR/MPL is tested, according to current European Society for Medical Oncology and National Comprehensive Cancer Network guidelines. Such testing is typically performed with allele specific assays for canonical mutations only, such as JAK2 Val617Phe and MPL Trp515Leu/Lys/Ala.

However, NGS allows for direct and concurrent analysis of these genes with high sensitivity.

In the analysis, investigators reviewed the JAK2, CALR, and MPL mutational profiles of patients with a suspected or diagnosed MPN who had been referred for molecular testing to identify mutational patterns that may arise in diagnostic misclassification.

Targeted NGS was prepared with a custom Access Array system (Fluidigm) encompassing 26 myeloid genes, including JAK2 exons 12, 14, 16, CALR exon 9, and MPL exon 10, and sequenced on a MiSeq (Illumina) to a minimum depth of 1000x. Single cell DNA libraries were prepared with the Tapestri platform (Mission Bio) and sequenced on a NextSeq (Illumina).

A total of 2664 patients with a suspected or diagnosed MPN underwent targeted NGS from 2014 to 2019. A total of 743 JAK2, 282 CALR, and 74 MPL mutations were identified in 740 (28%), 282 (11%), and 73 (3%) patients, respectively.

Targeted NGS identified 4 (0.5%) non-canonical JAK2 mutations that were not c.1849G>T (Val617Phe; n = 730) or an exon 12 mutation (n = 9). These 4 variants included c.1849_1851delinsTTT and c.1848_1849delinsCT mutations that result in a Val617Phe change.

Additionally, JAK2 exon 12 mutations were identified in a subgroup of patients with a clinical phenotype of ET without evidence of polycythemia. Twelve (16%) MPL mutations were non-canonical (ie. not Trp515Leu/Lys/Ala) and included Ser505Asn, Trp515Arg/Ser, Val501Ala and 2 in-frame insertions. With allele specific assessment, these patients risk misclassification as a triple-negative MPN or exclusion of MPN altogether.

“Erroneous exclusion of an MPN...may lead to the further and unnecessary investigation of secondary causes,” wrote the authors.

Moreover, co-mutation of the JAK2 V617F mutation was reported with a CALR (n = 6) or MPL (n = 7) mutation. In these instances, the variant allele frequency of the CALR and MPL mutation was significantly higher than the co-mutated JAK2 (CALR vs JAK2 mean variant allele frequency = 25% vs 0.8%; MPL vs JAK2 = 22% vs 7%).

“Sequential testing strategies risk misclassifying the genetic driver in this group as JAK2 despite this mutation typically representing a minority clone compared to the co-mutated CALR/MPL,” the authors wrote.

In order to better understand the clonal makeup of these cases, investigators performed custom amplicon single-cell sequencing of an additional 2 patients with JAK2/CALR or JAK2/MPL co-mutations. The mutations were identified in non-overlapping sub-clones, consistent with clonal independence.

Reference

Thompson E, Yeh P, Nguyen T, et al. Co-mutations and non-canonical driver mutations identified by next generation sequencing in myeloproliferative neoplasms reveal significant opportunities for improved diagnostic practice. Presented at: 2020 European Hematology Association Congress; June 11-21, 2020; Virtual. Abstract EP1087.

Related Videos
Akriti Jain, MD
Mikkael A. Sekeres, MD, MS
Somedeb Ball, MBBS, assistant professor, medicine, Division of Hematology Oncology, Department of Medicine, Vanderbilt University Medical Center
Mikkael A. Sekeres, MD, MS
Naseema Gangat, MBBS