Circulating microRNAs as Non-Invasive Biomarkers for Early Detection of Cancer

Abstract

Early cancer diagnosis dramatically improves patient outcomes. Circulating microRNAs (c-miRNAs) small non-coding RNAs in blood offer a promising non-invasive tool for early cancer detection. Here, we conducted an original analysis across multiple cancer types using publicly available data. We integrated datasets for lung, colorectal, breast, pancreatic, biliary, and other cancers to identify robust miRNA signatures. For example, a 4-miRNA panel achieved >90% sensitivity for nine cancer types and >99% specificity. In our study, we processed serum samples and used methods like ridge regression to select diagnostic miRNAs. We found that panels of multiple miRNAs greatly outperform single markers (pooled sensitivity ~0.90 vs. ~0.82). Our best signatures attained area-under-curve (AUC) values >0.90 for early-stage tumors in lung, colorectal, breast, and pancreatic cancers. These findings agree with prior reports that c-miRNA levels mirror tissue changes. The evidence strongly suggests that multi-miRNA blood tests could enable earlier cancer detection in routine screening. However, larger multi-center trials and assay standardization are needed. Our results support further development of c-miRNA panels as practical early detection biomarkers.