HemaSphere (Nov 2022)

Different Gene Sets Are Associated With Azacitidine Response In Vitro Versus in Myelodysplastic Syndrome Patients

  • Marguerite-Marie Le Pannérer,
  • Jeannine Diesch,
  • Raquel Casquero,
  • Michael Maher,
  • Olga Garcia,
  • Torsten Haferlach,
  • Johannes Zuber,
  • Andrea Kündgen,
  • Katharina S. Götze,
  • Marcus Buschbeck

DOI
https://doi.org/10.1097/HS9.0000000000000792
Journal volume & issue
Vol. 6, no. 11
p. e792

Abstract

Read online

Myelodysplastic syndromes (MDS) are a heterogeneous group of hematopoietic disorders characterized by dysplasia, ineffective hematopoiesis, and predisposition to secondary acute myeloid leukemias (sAML). Azacitidine (AZA) is the standard care for high-risk MDS patients not eligible for allogenic bone marrow transplantation. However, only half of the patients respond to AZA and eventually all patients relapse. Response-predicting biomarkers and combinatorial drugs targets enhancing therapy response and its duration are needed. Here, we have taken a dual approach. First, we have evaluated genes encoding chromatin regulators for their capacity to modulate AZA response. We were able to validate several genes, whose genetic inhibition affected the cellular AZA response, including 4 genes encoding components of Imitation SWItch chromatin remodeling complex pointing toward a specific function and co-vulnerability. Second, we have used a classical cohort analysis approach measuring the expression of a gene panel in bone marrow samples from 36 MDS patients subsequently receiving AZA. The gene panel included the identified AZA modulators, genes known to be involved in AZA metabolism and previously identified candidate modulators. In addition to confirming a number of previously made observations, we were able to identify several new associations, such as NSUN3 that correlated with increased overall survival. Taken together, we have identified a number of genes associated with AZA response in vitro and in patients. These groups of genes are largely nonoverlapping suggesting that different gene sets need to be exploited for the development of combinatorial drug targets and response-predicting biomarkers.