Identification of clinically useful plasma miRNA as minimally invasive biomarkers for early stage Non-Small Cell Lung Carcinoma (NSCLC)

Lung cancer is currently the leading cause of cancer death in both developing and developed countries. Given that lung cancer has poor prognosis in later stages, it is essential to achieve an early diagnosis to maximise a patient’s overall survival. Lung adenocarcinoma (LUAD), a subtype of non-small cell lung cancer (NSCLC), has the highest incidence in both smokers and non-smokers. Therefore, there is an urgent need to develop methods to provide both sensitive and specific lung cancer diagnosis.The current standard and non-invasive screening method, low‐dose computed tomography (LDCT), is the only radiological method which shows to have mortality benefits from multiple large randomized clinical trials (RCT). However, these RCTs also found LDCT to have a significant false positive rate of 96% which resulted in unnecessary invasive biopsies being performed. Due to the lack of sensitive and specific screening methods for the early detection of lung cancer, there is an urgent need for alternative biomarkers that are minimally invasive and may provide diagnostic, and/or prognostic potential. Circulating biomarkers can be readily detectable in blood and have been extensively studied as prognosis markers. Circulating microRNA (miRNA) been investigated for these purposes as an augmentation to LDCT, or as direct diagnosis of lung cancer. There is however, a lack of consensus across the studies on what miRNAs are the most clinically useful.   In steps to address this problem, we performed a comprehensive unbiased assessment of all known miRNAs in a progression series of 25 human LUAD tissues. In the microarray data, we identified four novel findings. Firstly, we identified 13 miRNAs with differential expression changes across all four stages of LUAD tissue. Secondly, these 13 miRNAs include miR-4454, miR-4634 miR-4707-5p, and miR-4492, which have not been reported in previous literature related to early LUAD. Thirdly, three of the 13 miRNAs: miR-96-5p, miR-182-5p, and miR-183-5p also show a strong correlation between LUAD disease progression and expression level. Finally, we identified miR-31-5p as the strongest correlation with LUAD stage progression out of all the miRNAs investigated in our dataset.We validated the microarray expression of miR-31-5p, miR-96-5p, miR-182-5p, and miR-183-5p expression level by quantitative real time polymerase chain reaction (qRT-PCR) in the 25 LUAD tissues present on the array. Next, to confirm that miRNA expression may be related early stage LUAD, we obtained an independent cohort of stage I LUAD (n=23) and 11 adjacent normal tissues. Real-time qPCR confirmed that all four miRNAs’ expression level were significantly upregulated (P≤0.05) in stage I LUAD with miR-31-5p increasing by ~35-fold, miR-96-5p by ~8-fold, miR-182-5p by ~5-fold, and miR-183-5p by ~6-fold.Next, to explore the notion that the miRNA panel may serve as circulating biomarkers for early detection and/or prognosis, we obtained plasma samples from 56 patients (32 Stage I, 5 Stage II, 8 Stage III, and 5 Stage IV) with LUAD along with six healthy control samples (this set includes 26 samples which had a matched  tissue sample). As previously observed in stage I tissue, the circulating expression levels of miR-31-5p, miR-182-5p, and miR-183-5p were significantly increased. Interestingly, miR-96-5p expression was only detected in three out of 56 plasma samples and was deemed not to be a robust biomarker. As such, only the diagnostic utility of miR-31-5p, miR-182-5p, and miR-183-5p were investigated further. The receiver operator curve (ROC) area under the curve (AUC) of miR-31-5p, miR-182-5p, miR-183-5p ranged from 0.8750 - 0.8229 when stage I was compared with normal controls. Similar AUC scores were observed when stage II tissues were investigated. Therefore, these three miRNAs may have diagnostic utility for early detection of LUAD in plasma which warrants further investigation.In summary, we have identified miR-4454, miR-4707-5p, miR-4492, and miR-4634 as novel miRNA related to LUAD from the microarray. Moreover, miR-31-5p, miR-96-5p, miR-182-5p, and miR-183-5p tissue expression level have a strong correlation with LUAD disease progression. Finally, plasma miR-31-5p, miR-182-5p, and miR-183-5p have non-invasive diagnostic utility individually in early staged LUAD.