Integration of Clinical Data With Diagnostic Genome Profiling May Produce Personalized Prognostic Predictions in MDS

Several distinct subgroups have been identified within the myelodysplastic syndrome genomic landscape and have notable clinical features and discrete evolution patterns that provide proof of concept for next-generation disease classification and prognosis.

Several distinct subgroups have been identified within the myelodysplastic syndrome (MDS) genomic landscape and have notable clinical features and discrete evolution patterns that provide proof of concept for next-generation disease classification and prognosis, according to results from a study (NCT04174547) published in the Journal of Clinical Oncology.1

Investigators utilized Dirichlet processes to identify genomic subgroups among patients with MDS and found 6 components; each component individually described specific distribution variables. By performing hierarchical agglomerative clustering, 8 groups, or clusters, were then identified and they were defined using specific genomic features.

The 8 genomic-based MDS groups that were identified were found to have different probability of survival. Groups 1 and 6, which were characterized by the presence of SF3B1 mutations, were found to have better survival than groups 2 to 5 and group 7 (P from <.0001 to .0093). Moreover, isolated SF3B1 mutations (group 6) were linked with better outcomes with respect to SF3B1, with co-mutated patterns (group 1, P = .0304). Group 0, which included those without specific genomic abnormalities, was also found to have a good prognosis vs groups 2 to 5 and 7 (P from < .0001 to .012).

Groups that were defined by splicing mutations other than SF3B1 demonstrated a worse survival. Group 5, which included patients with SRSF2 mutations and co-existing mutations in other genes, was found to be associated with dismal outcomes vs groups 0, 1, 4, and 6 (P from < 0.0001-0.0177). Group 2, which was comprised of patients who harbored TP53 mutations and complex karyotypes were found to experience the worst outcomes (P from < .0001 to .0473).

Group 7, which included patients with acute myeloid leukemia (AML)–like mutations, had a worse prognosis than groups 1, 3, and 6 (P < .0001), and were also found to have a high rate of leukemic evolution. Lastly, those who had an isolated 5q deletion, those without mutations, or those with single mutations, were found to have a better prognosis than patients who had 2 mutations or more or TP53 mutations (P = .0432).

“The integration of clinical data with diagnostic genome profiling in MDS may provide prognostic predictions that are personally tailored to individual patients,” Matteo Bersanelli, PhD, of the Department of Physics and Astronomy, of the University of Bologna, and colleagues, wrote in the paper. “Such information will empower the clinician and support the complex decision-making process in these patients.”

Currently, disease-related risk in MDS is measured by the International Prognostic Scoring System (IPSS) by factoring in the percentage of bone marrow blasts, the number of peripheral blood cytopenias, and presence of clonal cytogenetic abnormalities. A revised version of the IPSS (IPSS-R) was proposed in 2012, and 5 cytogenetic risk groups were introduced, as well as refined categories for bone marrow blasts and cytopenias.2 Both the IPSS and IPSS-R are important tools that are used to assist in the clinical decision-making process, although both scoring systems have flaws that could result in failure to capture reliable prognostic information at the individual patient level.

For this study, investigators set out to define a new genomic classification of the disease and to improve prognostic assessment on a more personalized level.

To do this, an international, retrospective cohort of 2043 patients who had been diagnosed with primary MDS in accordance with World Health Organization criteria were evaluated. Additionally, the study also included an independent cohort of 318 patients who were prospectively diagnosed at the Humanitas Research Hospital in Milan, Italy.

Upon diagnosis, a cytogenetic analysis was conducted by utilizing standard G-banding. Karyotypes were then classified per the International System for Cytogenetic Nomenclature Criteria. In total, 47 genes related to myeloid neoplasms were assessed by screening DNA from peripheral blood granulocytes or bone marrow mononuclear cells.

All analyses conducted in the study were performed on the EuroMDS patient cohort (n = 2043), and the Humanitas cohort was used to independently validate models of prognostication.

Of the patients who were included from the EuroMDS cohort, 59% (n = 1195) had a normal karyotype and 32% (n = 651) had chromosomal abnormalities. Additionally, investigators identified mutations in 45 of the 47 genes that were assessed. Eighty percent (n = 1630) of patients presented with 1 or more mutations, with a median of 2 mutations per patient (range, 1-17). Six genes were mutated in over 10% of patients, 5 additional genes were mutated in 5% to 10% of patients, and 36 mutated genes were present in less than 5% of patients.

Utilizing these methods, investigators were able to categorize patients into several groups based on their genomic features. Group 0 included patients with MDS who did not harbor specific genomic profiles; these patients were characterized to be younger, have isolated anemia, normal or reduced marrow cellularity, an absence of ring sideroblasts, and a low percentage of marrow blasts.

Patients in group 1 had SF3B1 mutations with co-existing mutations in other genes like ASXL1 and RUNX1; these patients were characterized by anemia linked with mild neutropenia and thrombocytopenia, multilineage dysplasia, and higher marrow blast percentage with respect to group 6, at 7% vs 2%, respectively (P <.0001).

Patients in group 2 primarily had TP53 mutations and/or a complex karyotype, with most having 2 or more cytopenias and excess blasts. Moreover, group 3 featured patients with SRSF2 and concomitant TET2 mutations; these patients are characterized to present with single cytopenia and a higher monocyte count vs the other groups (P < .0001).

The dominant characteristics of group 4 included U2AF1 mutations linked with 20q deletion and chromosome 7 abnormalities. Those within this group had higher rates of transfusion-dependent anemia vs other groups (P from .023 to < .0001). Group 5 was characterized by harboring SRSF2 mutations, as well as co-existing mutations in ASXL1, RUNX1, IDH2, and EZH2. Patients within this group presented with 2 or more cytopenias, multilineage dysplasia, and excess blasts at a median of 11%, which was significantly higher than what was seen in group 3 (P = .031).

Group 6 featured patients with ring sideroblasts and isolated SF3B1 mutations except for TET2, DNMT3A, and JAK/STAT pathway genes. These patients were characterized by having isolated anemia, normal or high platelet count, single or multilineage dysplasia, and low percentage of marrow blasts, with a median of 2%. Lastly, patients in group 7 had AML-like mutation patterns, including DNMT3A, NPM1, FLT3, IDH1, and RUNX1. This population is characterized by having 2 or more cytopenias, with a high rate of transfusion dependency, as well as excess blasts that ranged from 15% to 19% in most cases.

Additionally, investigators examined different groups of patients with MDS to determine whether the genomic features could provide insight into how different patients will respond to specific treatment approaches. Investigators looked at 424 cases of patients treated with allogeneic transplant and 221 cases of patients who received hypomethylating agents.

“With the limit to analyze a retrospective cohort of selected patients, MDS groups on the basis of genomic features do not identify different [probabilities] of survival after hypomethylating agents, whereas they are able to significantly stratify post-transplantation outcome,” the study authors wrote.

Groups 1, 6, 7, and 0 appeared to have better outcomes following transplantation, and groups 2 and 4 were were found to have a high rate of transplantation failure.


  1. Bersanelli M, Travaglino E, Meggendorfer M, et al. Classification and personalized prognostic assessment on the basis of clinical and genomic features in myelodysplastic syndromes. J Clin Oncol. 2020;39(11):1223-1233. doi:10.1200/JCO.20.01659
  2. Greenburg PL, Tuechler H, Schanz J, et al. Revised International Prognostic Scoring System for myelodysplastic syndromes. Blood. 2012;120(12):2454-2465. doi:10.1182/blood-2012-03-420489