# Single-Cell Multi-Omics Identifies Measurable Residual Disease Targets Among Myelodysplasia- and Clonal Hematopoiesis-Related Genes in Acute Myeloid Leukemia

## TL;DR

This study uses single-cell multi-omics to identify personalized MRD markers in AML patients who lack traditional markers, enabling earlier detection of relapse.

## Contribution

The study introduces a novel approach using single-cell multi-omics to identify MRD markers in AML patients without validated molecular variants.

## Key findings

- Single-cell analysis identified MRD markers in all six AML patients lacking conventional markers.
- Markers detected relapses earlier than conventional methods like flow cytometry and WT1 monitoring.
- Some markers originated from myelodysplasia- or clonal hematopoiesis-related genes.

## Abstract

Acute myeloid leukemia (AML) patients are monitored using markers reflecting residual leukemic burden during therapy and throughout post-treatment surveillance, enabling informed clinical decision-making. However, approximately 50% of AML patients lack validated molecular AML-defining variants for measurable residual disease (MRD) monitoring, due to difficulties distinguishing leukemic markers from preexisting molecular aberrations. We hypothesized that single-cell multi-omics analyses enable identification of useful MRD markers in these patients by delineating the subclonal landscape. By analyzing samples from six AML patients, we identified markers for all patients and confirmed their leukemia specificity in remission samples using droplet digital PCR (ddPCR) or error-corrected next-generation sequencing (EC-NGS), which confirmed leukemia specificity for all markers. Noteworthy, the single-cell-guided markers revealed low levels of residual disease and detected relapses earlier than conventional MRD approaches (flow cytometry and WT1 overexpression), highlighting the possibility of personalized disease monitoring for patients currently excluded from sensitive MRD assessments.

Background: In acute myeloid leukemia (AML), the most sensitive measurable residual disease (MRD) methods are single-gene approaches, but these are applicable only in ~60% of AML cases. Methods: We applied multi-omics single-cell analysis on diagnostic and first remission samples to identify leukemia-specific molecular markers for subsequent MRD monitoring in six AML patients lacking AML-defining variants. Results: Five selection criteria were defined to identify suitable MRD markers. Markers of primordial leukemic clones were identified by combining data from single-cell sequencing and immunophenotyping. Specific markers suitable for use in MRD follow-up were identified in 6/6 patients, in some cases in myelodysplasia-related genes and clonal hematopoiesis-related genes usually not recommended for use in MRD determinations. Patient-specific ddPCR (limits of detection: 0.06–0.0011%) or EC-NGS assays correlated with therapeutic responses: 0/4 markers displayed molecular relapses in three non-relapsing patients, contrary to 4/4 markers of three relapsing patients. Of these, 3/4 and 1/4 markers detected molecular relapses earlier than or simultaneous with conventional methods, respectively (−115 to −338 days). Conclusions: Our results demonstrate that single-cell subclonal mapping at diagnosis and during first remission enables selection of reliable MRD targets for personalized disease surveillance in patients lacking conventional MRD markers.