Related Experiment Video
Updated: Sep 19, 2025

Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances
Published on: October 11, 2018
Integrative Genetic and Transcriptomic Subtyping Improves Prognosis Prediction in B-Lineage Acute Lymphoblastic
Mercilena Benjamin1, Jay Singh1, Avanish Kumar Pandey1
1Laboratory Oncology, AIIMS, New Delhi, India.
None:
Whole-transcriptomic sequencing (WTS) has remarkably advanced our understanding of B-lineage acute lymphoblastic leukemia (B-ALL), allowing for detailed gene expression profiling and discovery of novel therapeutically relevant subtypes. The aim of this study was to evaluate the diagnostic and prognostic relevance of combining WTS with traditional genetic methods in risk-stratifying B-ALL. In a cohort of 394 patients (301 children and 93 adults), conventional techniques such as fluorescence in situ hybridization, cytogenetics, and reverse-transcription PCR identified sentinel chromosomal abnormalities like BCR::ABL1, TCF3::PBX1, ETV6::RUNX1, and KMT2A-R (rearranged), and ploidy status. WTS was performed on selected 257 patients to identify subtypes such as Ph-like, DUX4-R, PAX5-altered (PAX5-ALT), MEF2D-R, BCL2-R, UBTF-R, PAX5 P80R, NUTM1-R, ZNF384-R, ZNF384-like, ETV6::RUNX1-like, IKZF1 N159Y, and HLF-R. We used a multipronged strategy to identify the borderline subtypes such as Ph-like, PAX5-ALT, and CRLF2 (non-Ph-like), by integrating gene expression signatures using t-distributed stochastic neighbor embedding, subtype-defining mutations, gene fusions, and copy number assessments. Our integrated approach not only identifies prognostically relevant sentinel molecular subtypes but also increases subtype assignment in upto ∼95% of B-ALL patients. The pro-B immunophenotype was found to be more frequent in UBTF-R and MEF2D-R ALL. Ph-like ALL was associated with poor remission rates and higher minimal residual disease positivity, while DUX4-R showed favorable prognosis. We further categorized pediatric patients into 3 risk groups: favorable (hyperdiploid, ETV6::RUNX1, and DUX4-R), poor (BCR::ABL1, Ph-like, KMT2A-R, TCF3::PBX1, iAMP21, and hypodiploid), and intermediate (PAX5-ALT, PAX5 P80R, NUTM1-R, MEF2D-R, CRLF2 [non-Ph-like], UBTF-R, ZNF384-R, ZNF384-like, BCL2-R, IKZF1 N159Y, ETV6::RUNX1-like, and B-rest). Event-free survival and overall survival were significantly associated with this risk stratification. In adults, Ph-like ALL showed worse prognosis, particularly, in BCR::ABL1-negative ALL patients. Among the DUX4-R B-ALL, those with IKZF1 deletion had worse event-free survival and overall survival. We also identified several novel gene rearrangements in different subtypes of B-ALL. Our study demonstrated that integrating WTS with traditional methods provides a comprehensive, accurate, and cost-effective strategy for risk assessment and treatment planning for B-ALL.

