# Circulating microRNAs as Biomarkers of Brain Metastases in Lung Cancer: A Pilot Study

## TL;DR

This pilot study explores whether specific microRNAs in the blood can serve as biomarkers for brain metastases in lung cancer patients.

## Contribution

The study identifies differentially expressed miRNAs in lung cancer patients with brain metastases and explores their potential as biomarkers.

## Key findings

- Two miRNAs (miR-409-3p and miR-485-3p) were significantly downregulated in lung cancer patients with brain metastases compared to healthy controls.
- Eleven miRNAs showed consistent selection in feature-ranking analysis, with miR-363-3p, miR-210-3p, and miR-194-5p exhibiting the highest stability.
- Predictive modeling showed limited utility of the identified miRNAs for distinguishing patients with and without brain metastases.

## Abstract

Background/Objectives: There is an ongoing search for reliable biomarkers of lung cancer (LC) and its progression, including nervous system involvement. MicroRNAs (miRNAs) play a crucial role in the regulation of gene expression and represent a promising focus of investigation in this field. The aim of this study was to assess the profile of miRNA expression in patients diagnosed with lung cancer, with or without brain metastases. Methods: This study comprised 13 patients diagnosed with non-small cell lung cancer (mean age 64.8 years, 61.5% females): 6 with brain metastases (LC + BM) and 7 without them (LC), and a control group of 6 healthy volunteers (HC). The expression levels of 179 miRNAs were assessed and compared between the study groups using quantitative reverse-transcription PCR (qRT-PCR). Results: In LC + BM subgroup, two miRNAs were found to be downregulated in comparison with HC: miR-409-3p (logFC = −17.42, p = 0.029) and miR-485-3p (logFC = −17.30, p = 0.026). An exploratory, probe-based feature-ranking analysis identified eleven miRNAs that were repeatedly selected across the resampling runs: miR-363-3p, miR-210-3p, miR-194-5p, miR-409-3p, miR-22-3p, miR-2110, miR-326, miR-485-3p, miR-223-5p, miR-16-2-3p, and miR-139-5p. Among these, miR-363-3p, miR-210-3p, and miR-194-5p exhibited the highest empirical stability. Predictive modeling was subsequently evaluated using a fully nested cross-validation framework in which feature selection and model training were repeated within each training fold. Under this stringent evaluation, the classification performance was close to chance across all the evaluated algorithms, indicating a limited predictive utility of the identified miRNAs for distinguishing patients with and without brain metastases in the present dataset. Conclusions: Notable differences in miRNA expression profiles were revealed for the patients with brain metastases from lung cancer, suggesting the role of the selected miRNAs in cancer metastasis to the CNS. However, while our analysis provides exploratory insights, the findings should be interpreted with caution and require validation in larger, independent cohorts before any clinical or translational implications can be established.