Mayo Alliance Prognostic Model for Myelodysplastic Syndromes: Integration of Genetic and Clinical Information

Mayo Clin Proc. 2018 Oct;93(10):1363-1374. doi: 10.1016/j.mayocp.2018.04.013. Epub 2018 Jun 14.

Abstract

Objective: To develop a new risk model for primary myelodysplastic syndromes (MDS) that integrates information on mutations, karyotype, and clinical variables.

Patients and methods: Patients with World Health Organization-defined primary MDS seen at Mayo Clinic (MC) from December 28, 1994, through December 19, 2017, constituted the core study group. The National Taiwan University Hospital (NTUH) provided the validation cohort. Model performance, compared with the revised International Prognostic Scoring System, was assessed by Akaike information criterion and area under the curve estimates.

Results: The study group consisted of 685 molecularly annotated patients from MC (357) and NTUH (328). Multivariate analysis of the MC cohort identified monosomal karyotype (hazard ratio [HR], 5.2; 95% CI, 3.1-8.6), "non-MK abnormalities other than single/double del(5q)" (HR, 1.8; 95% CI, 1.3-2.6), RUNX1 (HR, 2.0; 95% CI, 1.2-3.1) and ASXL1 (HR, 1.7; 95% CI, 1.2-2.3) mutations, absence of SF3B1 mutations (HR, 1.6; 95% CI, 1.1-2.4), age greater than 70 years (HR, 2.2; 95% CI, 1.6-3.1), hemoglobin level less than 8 g/dL in women or less than 9 g/dL in men (HR, 2.3; 95% CI, 1.7-3.1), platelet count less than 75 × 109/L (HR, 1.5; 95% CI, 1.1-2.1), and 10% or more bone marrow blasts (HR, 1.7; 95% CI, 1.1-2.8) as predictors of inferior overall survival. Based on HR-weighted risk scores, a 4-tiered Mayo alliance prognostic model for MDS was devised: low (89 patients), intermediate-1 (104), intermediate-2 (95), and high (69); respective median survivals (5-year overall survival rates) were 85 (73%), 42 (34%), 22 (7%), and 9 months (0%). The Mayo alliance model was subsequently validated by using the external NTUH cohort and, compared with the revised International Prognostic Scoring System, displayed favorable Akaike information criterion (1865 vs 1943) and area under the curve (0.87 vs 0.76) values.

Conclusion: We propose a simple and contemporary risk model for MDS that is based on a limited set of genetic and clinical variables.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Age Factors
  • Aged
  • Bone Marrow Examination / statistics & numerical data
  • Core Binding Factor Alpha 2 Subunit / genetics
  • Female
  • Hemoglobins / analysis
  • Humans
  • Karyotyping / methods
  • Karyotyping / statistics & numerical data
  • Male
  • Mutation
  • Myelodysplastic Syndromes* / blood
  • Myelodysplastic Syndromes* / diagnosis
  • Myelodysplastic Syndromes* / genetics
  • Myelodysplastic Syndromes* / mortality
  • Phosphoproteins / genetics
  • Platelet Count / statistics & numerical data
  • Prognosis
  • Proportional Hazards Models
  • RNA Splicing Factors / genetics
  • Repressor Proteins / genetics
  • Reproducibility of Results
  • Risk Assessment / methods*
  • Risk Factors
  • Sex Factors

Substances

  • ASXL1 protein, human
  • Core Binding Factor Alpha 2 Subunit
  • Hemoglobins
  • Phosphoproteins
  • RNA Splicing Factors
  • RUNX1 protein, human
  • Repressor Proteins
  • SF3B1 protein, human