|
|
Background: Current prognostic tools for chronic myeloid leukemia (CML) rely primarily on clinical parameters and lack reliable molecular biomarkers for dynamic risk assessment. Circular RNAs (circRNAs) have emerged as promising liquid biopsy biomarkers because of their stability in peripheral blood. However, their prognostic value in CML remains underexplored.
Methods: In this retrospective cohort study, peripheral blood samples were collected from 417 patients newly diagnosed with CML and 45 healthy controls. Differentially expressed circRNAs were identified by high-throughput sequencing and validated by quantitative real-time PCR. The patients were randomly divided into training (n = 292) and validation (n = 125) cohorts. A prognostic nomogram incorporating circRNAs and clinical variables was developed using multivariate Cox regression, with model performance assessed using the area under the curve (AUC), calibration, and decision curve analysis.
Results: Two circRNAs (hsa_circ_0001523 and hsa_circ_0000095) were significantly dysregulated in patients with advanced-phase CML (p < 0.0001). The nomogram combining the expression patterns of two circRNAs with basophil%, blast%, and hemoglobin levels achieved an AUC of 0.920 (95% CI: 0.873 - 0.968) for survival prediction. High-risk patients (score > 122.5) had a significantly shorter median survival (22 vs. 83 months, p < 0.0001).
Conclusions: This circRNA-based nomogram is a clinically applicable, non-invasive tool for dynamic risk stratification in CML. The integration of molecular and hematological markers may guide personalized therapy decisions and address the limitations of existing prognostic systems.
DOI: 10.7754/Clin.Lab.2025.251027
|
