Work

Validation of blood-based detection of breast cancer highlights importance for cross-population validation

Year: 2025

Type: article

Abstract: Wang et al. 1 recently described a new PCR-based test for breast cancer (BC) detection based on DNA methylation (DNAme) in peripheral blood-derived monocytes (PBMCs), reporting a strikingly high area under the curve (AUC) of 0.94 when combining four DNAme sites in a multiplex-based methylation-specific quantitative PCR assay. Individual loci within their multiplex assay achieved an AUC of at least 0.69 in their discovery set derived from Illumina Methylation array data. In independent whole blood methylation datasets, we observe a limited ability to distinguish BC cases from controls using these loci ( n = 208 controls, n = 102 cases; maximum AUCs of 0.59 for individual, or 0.60 for combined loci, respectively), underscoring the importance of cross-population validation of diagnostic biomarkers prior to clinical implementation, although a predictor for an Asian population, even if it does not (or only weakly) validate in a European population, is still important. Notably, we show that individuals with systemic sclerosis and rheumatoid arthritis (RA) show similar changes in selected sites, suggesting that the observed signal may be at least in part associated with inflammation.

Source: Nature Communications

DOI: 10.1038/s41467-025-57265-z

Authors: Bente Theeuwes, Srikant Ambatipudi, Zdenko Herceg, Chiara Maria Herzog, Martin Widschwendter