Diagnostic Accuracy and Clinical Translation Challenges of Circulating MicroRNA Panels for Early-Stage Breast Cancer Detection: Meta-Analysis

Results: The four inclusion criteria were met in fourteen studies (2128 breast cancer cases and 2117 controls). The pooled diagnostic odds ratio (DOR) was 63.864 (95% CI: 28.820–141.516), with positive and negative likelihood ratios of 6.099 (95% CI: 3.906–9.525) and 0.101 (95% CI: 0.050–0.204), respectively. The performance of serum-based assays was better than plasma-based assays (DOR: 98.577 vs. 38.271). The best diagnostic accuracy was achieved by 3-miRNA panels (DOR: 126.599) and was saturated at more than 4 miRNAs panels. Pooled estimates were highly heterogeneous (I²>85%) in most cases, and this reduces the interpretation and generalizability of pooled estimates. Sensitivity analyses showed results were robust with no evidence of publication bias (p = 0.47). Conclusions: miRNA panels are proof of concept in case-control studies (sensitivity 91.5%, specificity 88.1%, DOR 63.864); serum miRNA panels (3-4 miRNAs) have better performance in case-control studies (serum DOR: 98.577 vs plasma: 38.271). However, there is significant heterogeneity (I²>85%) and case-control designs that reduce the applicability of the clinical. These results warrant further studies but cannot be used to demonstrate clinical feasibility. A validation of prospective screening is necessary.