surviBALL: exploring lncRNA expression at diagnosis for 5-year EFS risk stratification in pediatric B-ALL-a proof of concept.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
BACKGROUND: B-cell Acute Lymphoblastic Leukemia (B-ALL) remains an important cause of cancer-related death in children. Therefore, accurate identification at diagnosis of patients at high risk of relapse is crucial. In this context, long non-coding RNAs (lncRNAs) could be novel candidates with great potential. Hence, the aim of this study was to identify new prognostic biomarkers in pediatric B-ALL through an RNA sequencing (RNA-seq) approach that allows the detailed exploration of a wide range of lncRNAs. METHODS: Total RNA from two cohorts of B-ALL patients (C1 with 50 Spanish patients, and C2 with 72 Canadian patients) was sequenced with a depth of approximately 150 million paired-reads using Illumina technology. All protein coding and non-coding genes included in lncRNAKB annotation were studied to develop a gene expression-based 5-year Event Free Survival (EFS) prediction model. RESULTS: First, univariate Cox proportional hazards analyses identified 48 genes significantly associated with higher EFS risk in both cohorts. From these, ALASSO regression selected five genes, all of which are lncRNAs, as the most informative to develop the prediction model, which we have called surviBALL. Stratification of patients into three risk groups according to the surviBALL model revealed significantly poorer EFS in high-risk patients across C1, C2, and the integrated C1 + C2 cohort (P < 0.001). Validation in an independent cohort of 177 publicly available B-ALL samples confirmed surviBALL's prediction capacity (P = 2.80 × 10- 4) and its independence of both subtype and MRD. CONCLUSIONS: These findings suggest that surviBALL has the potential to complement current risk stratification approaches, particularly by identifying patients at high risk of relapse at diagnosis. As a hypothesis-generating proof of concept, this study highlights the promise of more personalized treatment strategies and warrants further validation in independent cohorts.