NEW YORK (GenomeWeb) – A group of Italian researchers have validated a microRNA-based blood test for early detection of lung cancer that they hope to introduce as a first-line tool that doctors can use to select which patients should go on to receive additional screening with low-dose computed tomography (LDCT).
The so-called miR-Test analyzes patients' serum based on a 13-miRNA signature and deems them either positive or negative for lung cancer. miRNAs are frequently abnormally expressed in tumors, and this irregularity shows up in miRNA profiles in serum and plasma.
In a recent article in the Journal of the National Cancer Institute, researchers led by Fabrizio Bianchi at the European Institute of Oncology (EIO) in Milan described the development and validation of the miR-Test involving 1,000 individuals previously enrolled in the Continuous Observation of Smoking Subjects (COSMOS) trial, who were 50 years of age or older and heavy or former smokers. In the JNCI paper, Bianchi and colleagues reported the miR-Test had a negative-predictive value of greater than 99 percent, but they are planning to further study it in a larger cohort before making it broadly available to patients.
In developing the test, Bianchi and his colleagues first narrowed the miRNA signature of interest from an initial set of 34 markers down to 13, by analyzing samples from 24 subjects. There was strong correlation between the two signatures, but the researchers homed in on a signature involving fewer markers to reduce costs and test complexity.
The researchers then validated the test in more than 1,000 COSMOS trial participants. The miRNA signature gave risk scores that were statistically significantly different between cancer patients and healthy subjects, with an accuracy of 74.9 percent, sensitivity of 77.8 percent, and specificity of 74.8 percent.
"The logic behind the development of miR-Test was to design a low-invasive blood test that can identify high-risk individuals who may benefit of LDCT screening for lung cancer early detection," Bianchi told GenomeWeb over e-mail. Researchers would like to advance the miR-Test as a tool that doctors can use before screening patients with LDCT. "We are not proposing a test to replace the LDCT exam," Bianchi noted.
At present time, doctors consider patients' age and smoking status as the two main factors when modeling their risk for lung cancer. If patients are deemed to be at high risk for lung cancer by these factors, doctors recommend them for LDCT screening – a costly intervention that sometimes flags benign lung nodules as possibly cancerous. Approximately 28 percent of high-risk individuals who receive LDCT exams get a positive result, because the test frequently picks up pulmonary nodules. But up to 97 percent of pulmonary nodules turn out to benign, Bianchi pointed out.
"Therefore, a diagnostic test to be used before LDCT that allows reducing the rate of LDCT-positive results in the absence of lung cancer, will be advantageous in terms of reduction of hospitalization costs, exposure to radiation, and psychological stress," he said. "The properties of our miR-Test, indeed, address these issues."
In 83 participants without lung cancer, Bianchi's team gauged whether the miR-Test results would be confounded by benign tumors, nodules, and other non-cancerous conditions. Overall, the test returned a negative result for 72 participants (86.7 percent). Within this population, 18 out of 24 patients with pneumonia, 15 out of 16 patients with chronic obstructive pulmonary disease, 34 out of 38 people with stable pulmonary nodules, and all 5 individuals with benign tumors received a negative miR-Test result.
The researchers also assessed the miR-Test on a third, independent cohort of 74 patients with more advanced lung cancer (Stage II to III disease), which showed the test had a sensitivity of 70.3 percent.
Within the overall COSMOS cohort, in which patients received LDCT, the miR-Test gave negative results to 820 out of 1,115 participants. Of those with negative results, 810 didn't have lung cancer, but 10 did. The overall sensitivity of the miR-test in the study was 79.2 percent, which translates to a false-negative rate of 20.8 percent. Additionally, Bianchi and colleagues reported the test had a negative-predictive value of more than 99 percent.
The miR-Test's sensitivity and NVP were "comparable with LDCT alone, indicating [it] could be used as a first-line screening tool," Bianchi and colleagues concluded in their paper.
In the National Cancer Institutes' National Lung Screening Trial – a study enrolling more than 50,000 former heavy smokers that compared LDCT and standard chest x-ray – LDCT had a sensitivity of 93.8 percent and specificity of 73.4 percent. The US Preventive Services Task Force used the data from NLST to recommend annual screening for lung cancer patients with LDCT in people 55 to 80 years old, who are current smokers or had a 30 pack-year smoking history (i.e. one pack/day for 30 years; half-pack/day for 60 years, etc.) in the past 15 years. Additionally, the Centers for Medicare & Medicaid Services last year also agreed to cover yearly LDCT screenings for the asymptomatic, heavy-smoker, or past-smoker population.
However, subsequent analysis of the NLST data published in JAMA by Duke University's Edward Patz and colleagues reported around 18 percent of lung cancers detected by LDCT in the study "appear to be indolent." The authors concluded that "overdiagnosis should be considered when describing the risks of LDCT screening for lung cancer."
Since in the real world the miR-Test is intended to be used alongside smoking status and age to identify people at high risk of lung cancer, the population that would receive the test would be enriched with those who have malignant lung nodules, Bianchi noted. "This will ultimately augment the number of lung tumors diagnosed despite a slight reduction in sensitivity compared to LDCT alone," he said.
A cost-effectiveness analysis published last year in the New England Journal of Medicine using NLST data estimated that LDCT would cost an additional $1,631 per person and $81,000 per quality-adjusted life year gained. The researchers of that analysis also noted that "modest changes" to their analysis could alter their findings and ultimately concluded that "the determination of whether screening outside of the trial will be cost effective will depend on how screening is implemented."
The developers of the miR-Test believe it can better direct high-cost screening tests to those that need it most and positively impact the cost effectiveness of lung cancer screening. If the miR-Test reaches the market, it will be priced at around $250, according to Bianchi, which is significantly less than the cost of LDCT and other follow-up interventions, such as biopsies and imaging tests.
A number of other research groups are also working on developing miRNA-based screening tests for lung cancer. For example, researchers at the University of Maryland last year published their validation of a sputum miRNA lung cancer test in a comparatively smaller number of patients. Bianchi noted that his group's test is among the few miRNA tests being developed that have been validated in the context of a large population screening trial for the diagnosis of asymptomatic lung cancer.
Based on the results of this initial validation effort, researchers are now assessing the miR-Test in the COSMOS II trial, which involves around 10,000 participants who will be screened by both the miR-Test and LDCT. "This is an important step toward miR-Test clinical application," Bianchi said. The first COSMOS trial was performed at EIO, "which might represent a limit for miR-Test validation," he observed, adding that the COSMOS II trial will involve eight different health institutions.
Additionally, the larger COSMOS II study will look at the cost-benefit analysis of the diagnostic in early lung cancer detection. The results for that trial will be available at the end of 2017.