Scientists identify drug-resistant bone marrow cells linked to poor outcomes in fast-growing leukemia

Study links T-ALL tumor subpopulations to treatment failure, maps healthy pediatric blood development, and identifies a potential therapy targeting drug-resistant cells.

Study: A multiomic atlas identifies a treatment-resistant, bone marrow progenitor-like cell population in T cell acute lymphoblastic leukemia. Image Credit: Matt_Images / Shutterstock.com

In a recent study published in Nature Cancer, researchers explore treatment resistance and relapse in T-cell acute lymphoblastic leukemia (T-ALL) to identify specific subpopulations of leukemia cells that could be linked to insensitivity to treatment and relapse.

Improving T-ALL survival rates

T-ALL is a pediatric cancer with limited effective treatment options for cases of relapse, which leads to poor survival rates in these patients. Recent advances in B-cell acute lymphoblastic leukemia through targeted therapies and genetic risk assessments have not translated to the treatment of T-ALL, which is partly due to an inadequate understanding of the biological risk factors for this cancer.

T-ALL relapse often involves treatment-resistant subpopulations, thus indicating that tumor heterogeneity may be a critical factor in these cases. Chemotherapy-resistant populations, particularly in early T-cell precursor-ALL, exhibit developmental arrest in progenitor states that may enhance resistance to treatment.

However, current strategies have not successfully addressed these challenges, thus necessitating robust frameworks for identifying subpopulations that contribute to treatment resistance and relapse.

About the study

In the present study, researchers employed single-cell multiomics to examine over 600,000 cells from 40 cases of T-ALL. Ribonucleic acid (RNA) and chromatin accessibility data were integrated to construct detailed developmental maps of leukemic and healthy hematopoietic cells.

Patient samples were obtained from diagnostic bone marrow and peripheral blood. These samples were processed using advanced sequencing techniques including single-cell for transposase-accessible chromatin sequencing (scATAC-seq) assays, as well as cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq) to capture transcriptomic and epigenomic profiles.

A comprehensive atlas of healthy pediatric hematopoiesis derived from bone marrow and thymus tissues were also analyzed to contextualize the findings. Leukemic cells were projected onto these reference maps to identify developmental arrest points and classify distinct cell subpopulations.

Drug sensitivity was assessed using in silico and in vitro screenings of targeted agents against BMP-like cells. Differential gene and chromatin analyses were also performed to identify unique molecular signatures associated with resistance. Additionally, functional validation involved patient-derived xenografts to test therapeutic vulnerabilities.

Statistical modeling was applied to evaluate the prognostic relevance of identified biomarkers across a large clinical cohort to link cellular phenotypes to treatment outcomes. Mutations in the neurogenic locus notch homolog protein 1 (NOTCH1) gene and its impacts were also investigated to understand their regulatory effects on cell fate and treatment response.

Study findings

A specific BMP-like leukemia cell population was primarily involved in treatment resistance in T-ALL. BMP-like cells, which were characterized by stem-like and myeloid lineage markers, were enriched in patients with a chemotherapy-resistant form T-ALL. These cells exhibited reduced T-lineage commitment and elevated resistance to standard therapies, including corticosteroids.

Using molecular signatures, BMP-like cells exhibited poor event-free and overall survival across multiple T-ALL subtypes. NOTCH1 mutations were found to influence these developmental arrest states, with dual NOTCH1 mutations driving cellular differentiation away from the BMP-like state and correlated with better outcomes.

Both in vitro and in silico analyses identified BMP-like cells as highly sensitive to apoptosis-inducing drugs like navitoclax and venetoclax, thus presenting a potential therapeutic approach. The use of patient-derived xenograft models confirmed the efficacy of these agents in targeting resistant cell populations.

The comparative analysis revealed similar BMP-like features in non-early T-cell precursor-ALL, which further confirms their role in broader treatment resistance. However, stratifying patients based on BMP-like gene expression improved risk prediction, thereby enabling the identification of high-risk individuals, even within traditionally low-risk groups.

Distinct molecular pathways, including transcriptional networks involving myocyte enhancer factor 2C (MEF2C) and B-cell lymphoma/leukemia 11B (BCL11B), were implicated in maintaining BMP-like states, thus providing insights into their persistence and adaptability.

Conclusions

The study findings reveal a treatment-resistant BMP-like cell population in T-ALL that was associated with poor clinical outcomes. By identifying their molecular characteristics and vulnerabilities, the current study provides crucial insights that may improve risk assessments and lead to the development of more targeted therapies. Furthermore, these findings indicate that therapeutic agents like venetoclax have the potential to be effective against BMP-like cells, paving the way for more effective interventions.