Non-coding RNAs in chronic lymphocytic leukemia: A systematic review and meta-analysis to decode the diagnostic potential.

[BACKGROUND] Chronic lymphocytic leukemia (CLL) comprises around 25-30 % of leukemia cases in the West. Emerging evidence underscores the role of non-coding RNAs (ncRNAs) like miRNAs, lncRNAs, and CircRNAs in CLL pathogenesis and regulation. The unique properties of ncRNAs have given the potential as non-invasive diagnostic biomarkers for CLL.

[METHODS] PubMed, Web of Science, Scopus, ProQuest, and Embase databases were searched (from inception up to January 2024) for studies addressing the correlation of ncRNA expression levels with diagnosis of CLL. The QUADAS-2 tool was employed to evaluate the risk of bias in the included studies. The GRADE approach evaluated the certainty of the evidence for diagnostic test accuracy.

[RESULTS] A total of 14 studies including 934 CLL patients were analyzed. Evaluations focused on miRNAs and CircRNAs due to sufficient primary data. For miRNAs, pooled sensitivity: 0.84, specificity: 0.98, positive likelihood ratio (PLR): 42.19, negative likelihood ratio (NLR): 0.16, diagnostic odds ratio (DOR): 260.14; area under the curve (AUC): 0.96. For CircRNAs, pooled sensitivity: 0.69, specificity: 0.77, PLR: 3.01, NLR: 0.40, DOR: 7.51, AUC: 0.80. GRADE assessments indicated very low certainty of evidence for miRNAs and low certainty for CircRNAs.

[CONCLUSION] Both miRNAs and CircRNAs appear to hold promise as non-invasive diagnostic biomarkers in CLL, with miRNAs demonstrating higher diagnostic performance. However, based on the GRADE assessments, the certainty of evidence may undermine the reliability of the pooled estimates. Future studies with rigorous design, larger sample sizes, and standardized protocols are essential to strengthen the certainty of evidence.