A Venetoclax/Azacytidine Single-Cell RNA-Seq Resistance Signature (VARS) Predicts Response and Relapse in AML Patients

• Published in: Blood 2024-11, Vol.144, p.5744-5744
• Main Authors: Merbach, Anne Kathrin, Rohde, Christian, Beneyto-Calabuig, Sergi, Stelmach, Patrick, Leppä, Aino-Maija, Waclawiczek, Alexander, Renders, Simon, Lotze, Michelle, Velten, Lars, Trumpp, Andreas, Sauer, Tim, Shahswar, Rabia, Heuser, Michael, Müller-Tidow, Carsten, Janssen, Maike
• Format: Article
• Language: English
• ISSN: 0006-4971
• DOI: 10.1182/blood-2024-209706

Acute myeloid leukemia (AML) is a highly aggressive and fatal disease, predominantly affecting older adults with a median diagnosis age of 68 years. Venetoclax, a BCL-2 inhibitor, combined with azacitidine (VA) has emerged as a standard treatment for AML patients ineligible for intensive chemotherapy. Despite initial efficacy, resistance to VA therapy remains a significant challenge, with approximately 27% of patients failing to respond and most responders eventually relapsing. Targeting anti-apoptotic proteins or combining venetoclax with IDH or FLT3 inhibitors may overcome therapy resistance and prolong relapse free survival. However, the limited success of these strategies in achieving long-term remission underscores the need to identify the relevant resistance cell populations and underlying mechanisms. We used our recently developed CloneTracer (Beneyto-Calabuig & Merbach et al., Cell Stem Cell) to investigate VA resistance in paired specimens. We performed single-cell multiomics analysis of 28 samples derived from 14 paired AML patients at the time of diagnosis and refractory/relapsed disease after VA. The multiomics dataset comprised single-cell transcriptomics, selected genomic and mitochondrial mutations, and for a subset of six patients, an immunophenotypic profile encompassing 82 surface markers. CloneTracer allowed intrapatient comparison between leukemic and healthy myeloid progenitors at similar stages of differentiation. To identify differentiation- and clone independent deregulated genes, we performed differential gene expression analysis for each cell type and clone of each patient, respectively. Subsequent sign analysis revealed a gene set consistently deregulated across multiple patients, cell types, and clones at relapse/refractory stages. We used this set of genes to calculate a score for each single cell, whereby the median score for each patient and cell type inversely correlated with relapse-free survival (RFS) at the time of diagnosis and thus represents a potential VA-resistance-score (VARS). This close association was most pronounced in the erythroid progenitor compartment. Moreover, comparison of the mean potential VA resistance score of refractory and relapsed patients, revealed that this score could identify patients with upfront refractoriness to VA, despite the heterogeneous genomic backgrounds of these individuals. Differences in VARS were predominant in the erythroid progenitor compartment, but also in the hematopoietic st