Breast cancer is the most frequently diagnosed malignancy among women worldwide and a leading cause of cancer-related mortality, and in Peru it remains the most common cancer in women, with a mortality rate of 11.0 per 100,000 inhabitants in 2021. Part of that toll reflects late detection. Mammographic screening, the backbone of early diagnosis, loses sensitivity in younger women and in those with dense breast tissue, and coverage across many Latin American countries remains suboptimal because of socioeconomic and geographical barriers. Those limitations are precisely why the field keeps searching for complementary, non-invasive molecular biomarkers that could be measured in a simple blood draw.

The biological rationale rests on microRNAs, small non-coding RNA molecules roughly 20 to 24 nucleotides long that regulate gene expression post-transcriptionally by binding target messenger RNAs and blocking their translation or promoting their degradation. Because miRNAs are secreted into the bloodstream, often packaged in vesicles or bound to carrier proteins, they are stable in plasma and can be quantified by reverse transcription quantitative PCR. The team selected four candidates, miR-21-5p, miR-486-3p, miR-451a, and miR-660-5p, based on a review of studies published between 2013 and 2024 that prioritized circulating miRNAs with reported diagnostic potential in breast cancer, particularly early-stage and Luminal disease. miR-21-5p has been consistently associated with breast cancer detection and was included in a diagnostic signature for early-stage disease, while miR-486-3p had been reported as significantly upregulated in Luminal A tumors compared with healthy controls, hinting at subtype-specific value.

The study design was an observational case-control analysis built around early-stage disease. The researchers recruited 36 control subjects without a family history of breast cancer, 32 patients with stage I or II Luminal A tumors, and 43 patients with stage I or II Luminal B tumors, all from the National Institute of Neoplastic Diseases (INEN) in Lima and all before any treatment. Every diagnosis was confirmed histologically and immunohistochemically. Under the INEN criteria, Luminal A tumors were defined as estrogen receptor- and progesterone receptor-positive, HER2-negative tumors with low Ki-67 proliferation below 15 percent, whereas Luminal B tumors were ER-positive with any PR status, either HER2 status, and high Ki-67 at or above 15 percent. All participants gave written informed consent, and the protocol was approved by the relevant ethics committees.

The laboratory workflow reflects the technical demands of circulating miRNA work. Peripheral blood was collected in K2-EDTA tubes, transported under a maintained cold chain at 4 degrees Celsius, and plasma was separated within three hours by sequential centrifugation, first at 1800g for 10 minutes and then at 20000g for 10 minutes at 4 degrees to strip away residual cellular debris. Aliquots were stored at minus 80 degrees and thawed only once before total RNA extraction using a commercial plasma RNA purification kit. Because miRNAs lack a poly(A) tail, the team used a polyadenylation and reverse transcription protocol that appends synthetic tails to the molecules, allowing primer-based cDNA synthesis, followed by miRNA-specific qPCR in technical duplicates with melt-curve analysis to confirm amplification specificity. RNU6 served as the endogenous reference for normalization, and fold-changes were calculated with the standard 2^(-ΔΔCt) method.

The expression results were strikingly asymmetric. Among the four candidates, miR-21-5p was significantly downregulated in Luminal A patients compared with controls, showing a 7.4-fold decrease, a direction contrary to most published reports of miR-21-5p overexpression in breast cancer. miR-486-3p, in contrast, was upregulated in Luminal A compared with both controls and Luminal B patients. Neither miR-451a nor miR-660-5p differed significantly across groups. The stability of miR-451a is itself informative, since it is highly abundant in erythrocytes and its elevation typically signals hemolysis; its flat profile across groups suggests that red-blood-cell contamination was minimal and did not distort the quantification of the other miRNAs.

To translate expression differences into diagnostic performance, the team built receiver operating characteristic models and, critically, validated them internally with leave-one-out cross-validation, a conservative approach in which each model is trained on all observations except one, iterated across the entire dataset. For distinguishing Luminal A from controls, miR-486-3p achieved an apparent AUC of 0.717, but after cross-validation the AUC settled at 0.651 and statistical significance was lost. For Luminal B versus controls, the combination of both miRNAs reached a validated AUC of 0.635. Most intriguingly, for the Luminal A versus Luminal B comparison, the two-miRNA model reached an apparent AUC of 0.790 with a significant p-value, which dropped to a validated AUC of 0.690 after cross-validation. These moderate values indicate real but limited discriminatory capacity, and the consistent shrinkage after validation is exactly what internal validation is designed to reveal, trimming the optimism bias that plagues small exploratory biomarker studies.

The researchers then asked whether the circulating miRNAs could add value to conventional clinical markers. Adding Ki-67 to the miR-486-3p and miR-21-5p model raised the validated AUC for Luminal A versus Luminal B discrimination from 0.690 to 0.785, and the three-marker model performed comparably to the classical immunohistochemical panel of ER, PR, HER2, and Ki-67, which achieved an AUC of 0.750. Ki-67 alone, however, outperformed everything with an AUC of 0.855. That benchmark matters because Ki-67 scoring is known to suffer from inter-laboratory variability and the lack of standardized cutoffs, so a circulating complement to proliferation scoring remains an attractive goal even if the miRNAs did not surpass it in this cohort.

Correlation and pathway analyses added biological texture. Spearman analysis revealed a strong positive correlation between miR-451a and miR-21-5p across the Luminal A and control groups, hinting at potential co-regulation, possibly through shared involvement in Wnt/beta-catenin signaling. Using the multiMiR database, the team identified 135 experimentally validated gene targets for miR-21-5p and six for miR-486-3p, with no shared targets between them. Exploratory KEGG enrichment showed that miR-21-5p targets were concentrated in cancer and inflammation-related pathways, including MiRNAs in cancer, MAPK signaling, and TNF signaling, whereas miR-486-3p targets were enriched in only two immune-related pathways, Fc gamma R-mediated phagocytosis and C-type lectin receptor signaling. The authors stress these analyses are hypothesis-generating, not mechanistic proof, but they suggest the two miRNAs may operate in complementary proliferative and immune processes in Luminal breast cancer.

The authors are candid about limitations. The sample size of 111 women limits statistical power, particularly for the null results on miR-451a and miR-660-5p, and the study lacked prospective hemolysis controls and no-RT controls for genomic DNA contamination. The unexpected downregulation of miR-21-5p in Luminal A patients may reflect molecular features of early-stage ER-positive tumors, since estradiol has been shown to suppress miR-21-5p expression in ER-positive contexts, underscoring why regional validation matters, as molecular signatures can vary across populations. Still, the overall message is one of cautious promise. Circulating miR-486-3p and miR-21-5p showed preliminary potential as candidate biomarkers for early-stage Luminal breast cancer, with miR-486-3p also hinting at subtype discrimination, and their integration with conventional clinicopathological markers could support future minimally invasive strategies for early detection and classification. Larger, independent cohorts are required before any clinical implementation, but for a population where mammography coverage is uneven and late detection is common, even a modest blood-based signal is worth chasing.

Subject of Research: Exploratory Analysis of Circulating miR‐486‐3p and miR‐21‐5p in Early Detection and Subtype Classification of Luminal A and B Breast Cancer in Peruvian Women

Article Title: Exploratory Analysis of Circulating miR‐486‐3p and miR‐21‐5p in Early Detection and Subtype Classification of Luminal A and B Breast Cancer in Peruvian Women

Article References: Bouroncle, D., Colque, G., Huaytalla‐Bautista, L., Flores‐Berrospi, M. C., Bernal, C., Belzusarri, I., Dunstan, J., & Paredes‐Anaya, M. (2026). Exploratory Analysis of Circulating miR ‐486‐3p and miR ‐21‐5p in Early Detection and Subtype Classification of Luminal A and B Breast Cancer in Peruvian Women. Cancer Reports, 9(10), Article e70717. https://doi.org/10.1002/cnr2.70717

Image Credits: AI Generated

DOI: 10.1002/cnr2.70717

Keywords: Exploratory, Analysis, Circulating, Early, Detection, Subtype, Classification, Luminal, Breast, Cancer, Peruvian, Women