NEW YORK (GenomeWeb) – Amid ever-increasing interest in using microRNAs as biomarkers, researchers from the University of Maryland this week published a new study suggesting that the small, non-coding RNAs may help in differentiating patients with early-stage lung cancer from those with non-malignant lung nodules.
Non-small cell lung cancer (NSCLC) remains the leading cause of cancer deaths in the US, killing an estimated 160,000 Americans last year. As early treatment can significantly improve patient outcomes, heavy smokers and people with other risk factors are routinely screened for the disease using low-dose computed tomography. Though effective, CT screening is expensive, time-consuming, and, importantly, often results in over-diagnosis since it frequently reveals non-cancerous lesions, or solitary pulmonary nodules (SPNs), as well as malignant ones.
"Right now, the challenge we face is how to identify real lung cancer [after] a positive discovery … by CT," Feng Jiang, the senior author of the report, told GenomeWeb. "Not every positive found … is lung cancer," which can lead to unnecessary procedures and patient anxiety.
"Therefore, we're trying to develop a non-invasive technique … to help identify lung cancer in nodules [identified as cancer] in CT," he added.
To do so, he is looking to sputum, which contains exfoliated bronchial epithelial cells and is easily obtained via bronchoscopy, and the miRNAs contained within.
In a previous study, Jiang and his colleagues profiled the expression of 818 human miRNAs in the sputum of 23 patients with NSCLC and 17 cancer-free individuals. They found that not only were the miRNAs stable and resistant to freezing and thawing, but could be sensitively and specifically detected by real-time RT-PCR. As described in a paper published in 2010, the researchers also identified 10 sputum-derived miRNAs whose expression was dysregulated in NSCLC patients.
Based on these findings and work of other groups, the University of Maryland team identified 13 sputum miRNAs — miR-21, -31, -126, -143, -155, -182, -200b, -205, -210, -372, -375, -486, and -708 — that show promise as biomarkers for NSCLC.
To build on these data, the researchers first sought to validate the miRNA set by analyzing sputum of 122 patients with either malignant or benign SPNs, according to a paper appearing this week inClinical Cancer Research.
In all patients, they were able to reliably detect the miRNAs in the sputum specimens using RT-qPCR, with the expression of each miRNA differing significantly between lung cancer patients and those with benign SPNs.
Through further analysis, the scientists winnowed down the miRNA set to three — miR-21, -31, and -210, all of which have been associated with various malignancies by others — that were significantly higher in patients who had NSCLC versus those who did not. When used in combination, the three miRNAs could differentiate between the patient types with 82.93 percent sensitivity and 87.84 percent specificity.
Notably, sputum cytology in these patients showed a sensitivity of 43.33 percent and a specificity of 90.32 percent. "Therefore, the sputum biomarkers had a higher sensitivity compared with sputum cytology, while maintaining a similar specificity," the researchers wrote in their paper.
In terms of the two primary forms of NSCLC, adenocarcinoma (AC) and squamous cell carcinoma (SCC), the expression of miR-21 was more closely associated with former, while miR-210 was related to the latter.
"However, overall, the panel of three biomarkers didn't exhibit special association with a particular histological type of the NSCLC cases," according to the Clinical Cancer Research paper. "Moreover, there was no association of the changes of the three genes with the age, gender, and ethnicity of the participants," although the size of the SPNs did influence miRNA signatures.
Encouraged by these findings, Jiang and his team next examined the three-miRNA panel in a blinded study with in-house sputum samples from 67 lung cancer patients and 69 individuals with benign SPNs. They found that this panel had 82.09 percent sensitivity and 88.41 percent specificity in differentiating malignant and benign SPNs, with no statistical differences between the stages of lung tumors.
The investigators also tested their panel in sputum samples from 76 lung cancer patients and 79 people with benign SPNs that were obtained from a different medical center. In this study, the miRNAs could be used to differentiate malignant from benign disease with about 81 percent sensitivity and 86 percent specificity.
However, while these numbers are encouraging, they are not sufficient to warrant the test's use in clinical settings, Jiang said. As such, he and his team are currently testing an expanded panel of miRNAs, using the three from the current panel and a number of others that have since been identified using newly unavailable next-generation sequencing techniques, with the goal of improving the test's sensitivity and specificity.
They are also aiming to test these additional miRNAs in a greater number of patient samples obtained from multiple clinics, he said.