• HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
Tuesday, August 11, 2026
BIOENGINEER.ORG
No Result
View All Result
  • Login
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
No Result
View All Result
Bioengineer.org
No Result
View All Result
Home NEWS Science News Cancer

Gene-expression risk model personalizes stem cell transplant decisions for children with AML

Bioengineer by Bioengineer
August 11, 2026
in Cancer
Reading Time: 4 mins read
0
Share on FacebookShare on TwitterShare on LinkedinShare on RedditShare on Telegram

Pediatric acute myeloid leukemia (AML) is one of the most aggressive childhood blood cancers, and the decision to use allogeneic hematopoietic stem cell transplantation (allo-HSCT) can profoundly influence a patient’s chances of survival. The procedure replaces diseased blood-forming cells with healthy donor stem cells, but it is also associated with serious risks, including graft-versus-host disease, severe infections, organ damage, infertility, and treatment-related mortality. Clinicians therefore face a difficult question at diagnosis: which children are likely to benefit from transplantation, and which may be exposed to unnecessary harm? A new study published in Genes & Diseases describes a transcriptome-based machine-learning framework designed to provide a more precise answer.

Researchers from Chongqing Medical University, Sun Yat-Sen University, and Foshan University developed a system called HSCT-64, which uses RNA-sequencing data to estimate both disease risk and the potential benefit of transplantation. Unlike conventional approaches that often depend heavily on minimal residual disease (MRD) measurements, HSCT-64 evaluates patterns of gene activity across a patient’s leukemia cells. These patterns, known as a transcriptomic profile, reflect the biological programs active within the cancer and can reveal differences in proliferation, differentiation, DNA repair, immune interaction, and resistance to treatment.

The framework contains two parallel prediction models. The first, known as aHSCT-64, is designed for patients who undergo allo-HSCT. The second, nHSCT-64, is intended for patients who do not receive transplantation. Rather than producing a single universal risk score, the two models generate separate prognostic rankings. By comparing where the same patient is ranked under each model, the researchers sought to identify whether transplantation was associated with a more favorable predicted outcome for that individual.

To build the system, the investigators combined four publicly available pediatric AML cohorts, creating a discovery dataset of 1,647 cases. They used machine-learning techniques to select and integrate 64 transcriptomic features associated with clinical outcomes. The resulting framework was then tested in independent pediatric AML datasets, including a separate Chinese cohort. This external validation step was essential because models developed from a single institution or population can perform well during initial testing yet lose accuracy when applied to patients treated elsewhere or profiled using different laboratory workflows.

The analysis focused on changes in risk ranking between the two parallel models. Patients whose predicted risk rank decreased when moving from the non-transplant model to the transplant model were classified as belonging to a potentially HSCT-benefiting subgroup. In the study’s framework illustration, these patients are represented by blue dots. By contrast, patients whose risk rank increased under the transplant model were designated as an HSCT-nonbenefiting subgroup, represented by red dots. The approach does not merely ask whether a patient has high-risk leukemia; it asks whether the expected outcome appears to improve specifically in the context of transplantation.

This distinction could be clinically important because high disease risk alone does not automatically mean that allo-HSCT will provide a net benefit. Some patients with aggressive disease may gain a survival advantage from the graft-versus-leukemia effect, in which donor immune cells attack malignant cells. Others may not respond sufficiently to transplantation or may face complications that outweigh its potential benefit. A model capable of separating overall prognosis from treatment-specific benefit could therefore support more individualized decisions than a strategy based on risk classification alone.

The researchers report that HSCT-64 maintained predictive performance across independent clinical datasets and identified groups with different apparent relationships between transplantation and outcome. Because the framework relies on RNA-sequencing measurements rather than a particular MRD platform, it may reduce some of the technical variability associated with laboratory-based residual disease testing. MRD remains a valuable component of pediatric AML care, but its interpretation can be influenced by the detection method, the molecular target available in a given patient, sample quality, timing, and differences in clinical judgment. A standardized molecular framework could provide an additional layer of evidence at diagnosis and during treatment planning.

The biological foundation of HSCT-64 is the idea that leukemia is not a uniform disease. Two children may receive the same diagnosis yet carry malignant cells with very different transcriptional programs. One tumor may be dominated by stem-like features and treatment resistance, while another may retain greater differentiation potential or interact differently with the immune system. RNA sequencing captures these genome-wide expression patterns, and machine learning can identify combinations that would be difficult to recognize using isolated biomarkers. The 64-feature signature is consequently intended to function as an integrated molecular readout rather than as a single-gene test.

The study does not suggest that HSCT-64 should immediately replace physician assessment, MRD testing, cytogenetic analysis, or established clinical risk factors. Instead, it presents the framework as a potential decision-support tool that could help identify children most likely to gain from transplantation while protecting others from avoidable toxicity. Before routine implementation, the model will require prospective evaluation, testing across additional international populations, and assessment of how it performs when RNA sequencing is conducted on different instruments and under different clinical conditions. Even so, the findings illustrate how transcriptomic medicine may move pediatric AML treatment beyond broad risk categories toward therapy selection tailored to the biology of each child’s leukemia.

Subject of Research:
A transcriptomic machine-learning framework for predicting allo-HSCT outcomes and identifying pediatric AML patients likely to benefit from transplantation.

Article Title:
A parallel-risk framework accurately predicts hematopoietic stem cell transplantation outcomes and identifies benefiting patients in pediatric AML

Web References:
https://doi.org/10.1016/j.gendis.2025.102003

References:
Genes & Diseases. DOI: 10.1016/j.gendis.2025.102003

Image Credits:
Yance Feng, Yali Shen, Ke Huang, Qian Li, Yu Tao, Rongqiu Liu, Liping Zhan, Hua Yang, Yang Xun, Yichao Xu, Wenli Tang, Binjun Xiong, Hui Shi, Liting Cheng, Li Wei, and Hua You

Keywords:
Pediatric acute myeloid leukemia; allo-HSCT; hematopoietic stem cell transplantation; RNA sequencing; transcriptomics; machine learning; precision oncology; minimal residual disease; HSCT-64; risk stratification

Tags: gene expression analysis for treatment planninggene-expression risk model for childhood leukemiagene-expression signatures predicting transplant benefitHSCT-64 gene activity profiling in AMLmachine learning models for pediatric cancer prognosispediatric AML stem cell transplant decision-makingpersonalized medicine in pediatric blood cancerspersonalized treatment strategies for pediatric AMLreducing transplant-related risks in childhood leukemiaRNA-sequencing for transplant risk assessmenttranscriptome-based machine learning in pediatric cancertranscriptomic profiling in acute myeloid leukemia

Share12Tweet7Share2ShareShareShare1

Related Posts

Recurrence-linked eccDNA patterns offer new prognostic insights for lung adenocarcinoma

August 11, 2026

Spatial Transcriptomics Reveals Prostate Cancer’s Molecular Evolution

August 11, 2026

Johns Hopkins Experts Co-Author New Treatment Framework for Liver Cancer

August 11, 2026

Biomarkers Improve Cost-Benefit Analysis of Head and Neck Cancer Immunotherapy

August 11, 2026

POPULAR NEWS

  • Gene Family Reveals How Tropical Crops Adapt to Heat

    29 shares
    Share 12 Tweet 7
  • Fungal Necrotizing Otitis Externa: Three Key Clinical Questions Answered

    29 shares
    Share 12 Tweet 7
  • More Electrons, Fewer Interfaces: Halide Cathodes Raise All-Solid-State Battery Energy Density

    29 shares
    Share 12 Tweet 7
  • Recurrence-linked eccDNA patterns offer new prognostic insights for lung adenocarcinoma

    29 shares
    Share 12 Tweet 7

About

We bring you the latest biotechnology news from best research centers and universities around the world. Check our website.

Follow us

Recent News

Gene Family Reveals How Tropical Crops Adapt to Heat

Fungal Necrotizing Otitis Externa: Three Key Clinical Questions Answered

More Electrons, Fewer Interfaces: Halide Cathodes Raise All-Solid-State Battery Energy Density

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 86 other subscribers
  • Contact Us

Bioengineer.org © Copyright 2023 All Rights Reserved.

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
No Result
View All Result
  • Homepages
    • Home Page 1
    • Home Page 2
  • News
  • National
  • Business
  • Health
  • Lifestyle
  • Science

Bioengineer.org © Copyright 2023 All Rights Reserved.