Gene panel testing for MDS/MPN overlap

A diagnosis of myelodysplastic syndromes/myeloproliferative neoplasms (MDS/MPN) overlap is given when a patient exhibits morphological features of both dysplasia and myeloid proliferation. However, precision in diagnosis is difficult owing to the heterogeneity of these clonal hematopoiesis disorders. As such, the identification of chromosomal and clonal abnormalities may help to differentiate diagnosis and improve treatment individualization.

In a recent review by Anoop K. Enjeti et al.,¹ published in Pathology, experts in MDS and MPN provided a recommended gene panel to be integrated into the diagnosis and prognosis of MDS/MPN clonal disorders. We summarize the key guidance below.

Recommended approach to using gene panel testing for MDS/MPN

In adults, there are five main subtypes of MPN/MDS overlap: atypical chronic myeloid leukemia (aCML); chronic myelomonocytic leukemia (CMML); MDS/MPN with ring sideroblasts and thrombocytosis (MDS/MPN-RS-T); MDS/MPN-unclassifiable (MDS/MPN-U); and therapy-related MDS/MPN (t-MDS/MPN). Incorporation of genetic testing in these disorders is used for risk stratification/prognostic scoring systems, and the most relevant mutations are shown in Table 1.

Table 1. Somatic mutations relevant to diagnosis and prognosis of MPN/MDS overlap*

aCML, atypical chronic myeloid leukemia; CMML, chronic myelomonocytic leukemia; CPSS-Mol, CMML Prognostic Scoring System-Molecular; MDS, myelodysplastic syndromes; MPN, myeloproliferative neoplasms; MDS/MPN-RS-T, MDS/MPN with ring sideroblasts and thrombocytosis; MDS/MPN-U, MDS/MPN-unclassifiable. *Table adapted from Enjeti et al.1
Overlap syndrome	Somatic mutation	Somatic mutations with prognostic relevance or included in risk stratification/prognostic scoring systems
CMML	TET2 (~60%) SRSF2 (~50%) ASXL1 (~40%) SETBP1 Concurrent TET2/SRSF2/RUNX1 (10–20%) TP53 (~1%)	Mayo molecular model (ASXL1 frameshift or nonsense mutation)               CPSS-Mol (ASXL1, NRAS or SETBP1 and RUNX1 mutations)               Groupe Francophone des Myélodysplasies (ASXL1 mutation)
aCML	ASXL1 (60%) SETBP1 (~20–40%) ETNK1 (~8%) NRAS/KRAS (35%)	Mayo prognostic model for aCML (TET2 mutation)               SETBP1
MDS/MPN-RS-T	SF3B1 (~90%) JAK2 V617F (50%) CALR (<5%) MPL (<5%) DNMT3A (~10–30%) SETBP1 (~10%)	Mayo prognostic model for MDS/MPN-RS-T (ASXL1 or SETBP1 mutations)
MDS/MPN-U	ASXL1 (~30–55%) TET2/SRSF2/JAK2 (~25%) SETBP1/EZH2/NRAS/KRAS (~10–15%)	ASXL1 is associated with poor prognosis               Overall prognosis is more dependent on cytogenetic and molecular features of individual cases

Certain mutational combinations are more common with certain subtypes of MDS/MPN, and when clonal features are combined with morphological identifiers, a more accurate picture of the subtype and disease course may be captured. The expert panel, led by chairs of the MDS, MPN, and working parties of the Australasian Leukaemia & Lymphoma Group (ALLG), provided recommended guidance for diagnosis and prognosis of MDS/MPN. Notably, gene testing does not drive differentiation, but rather acts as a supportive role in addition to these features for both diagnosis and prognosis (Figure 1). The authors also highlighted the possibility of using machine learning approaches when combining genotype-phenotype associations.

aCML, atypical chronic myeloid leukemia; AML, acute myeloid leukemia; CMML, chronic myelomonocytic leukemia; MDS, myelodysplastic syndromes; MDS/MPN-R-T, MDS/MPN with ring sideroblasts and thrombocytosis; MDS/MPN-U, MDS/MPN-unclassifiable; MPN, myeloproliferative neoplasms; T-MDS/MPN, therapy-related MDS/MPN.
*Adapted from Enjeti et al.^1
†Consensus panel of genes considered supportive in diagnosis, alongside morphological features of MDS/MPN: SF3B1, TET2, SRSF2, ASXL1, JAK2, NRAS, KRAS, CBL, RUNX1, SETBP1, MPL, CALR, ETNK1, DNMT3A, IDH1^‡, IDH2; and genes with prognostic relevance: TP53, EZH2, U2AF1.
^‡Rare in MDS/MPN.

Conclusion

The recommended gene panel for integration in the diagnosis and prognosis of MDS/MPN demonstrates the substantial heterogeneity of these disorders. Using this large data analysis approach, more individualized therapeutic guidance may be possible in MDS/MPN.