NEW YORK (GenomeWeb) – A panel of 19 microRNAs can distinguish malignant and benign thyroid nodules, which may help identify patients who would benefit from treatment.
While most thyroid nodules are benign, about 7 percent are malignant. Currently, suspicious thyroid nodules are evaluated by ultrasound-guided fine-needle aspiration biopsy and while that can settle some of the cases, about 30 percent of patients receive ambiguous results.
A Hadassah-Hebrew University Medical Center-led team of researchers analyzed 274 fine-needle aspiration biopsy samples to uncover a collection of miRNAs whose expression differs between benign and malignant thyroid nodules. As the researchers reported today in Cancer Epidemiology, Biomarkers, & Prevention, they found this miRNA panel to have high sensitivity and specificity in distinguishing benign and malignant thyroid nodules.
"Thyroid nodules are extremely common and occasionally harbor cancer," first author Haggi Mazeh from Hadassah-Hebrew said in a statement. "Patients who undergo evaluation of thyroid nodules often receive ambiguous results, and there are currently limited options for further management, representing a clinically unmet need."
The researchers recruited 137 adult patients who were undergoing thyroidectomies and obtained ex vivo fine-needle aspiration biopsies from their nodules and surrounding normal tissue. Of these 340 biopsies, 195 were classified as benign, 79 as malignant, and 66 were indeterminate.
At the same time, the researchers extracted RNA from these biopsy samples for deep sequencing on the Illumina HiSeq 2500 platform.
Examining a discovery group comprising 274 samples from 102 patients for whom a diagnosis was available, the researchers uncovered 19 miRNAs that were either under- or overexpressed in malignant samples, as compared to normal samples. They then combined these miRNAs into a prediction tool.
Many of these miRNAs, the researchers noted, have previously been implicated in cancer. For instance, miR-221/222 and miR-146b are also overexpressed in colon, breast, ovarian, hepatic, prostate, and other cancers.
A number of the miRNAs are thought to support the proliferation, migration, and invasion of thyroid cancer, though others are thought to inhibit its spread and lead to apoptosis, Mazeh added.
In their validation set of 66 samples from 35 patients, the prediction tool based on the panel of 19 miRNAs found 22 had malignant nodules, while 13 had benign ones. As compared to final histopathology, the panel had 91 percent sensitivity, 100 percent specificity, 87 percent negative predictive value, 100 percent positive predictive value, and 94 percent overall accuracy.
The researchers noted that their panel appears to perform better than current commercial molecular tests, which cost between $2,000 and $3,000. Those tests either have high sensitivities or high specificities, but not both, they said, making them only useful as "rule-out" tests. Mazeh and his colleagues noted that their panel could overcome those limitations.
"Our results provide a refined list of candidate microRNAs for diagnostic use, and confirms the perception that microRNA quantification is a promising form of molecular pathology," Mazeh said. "Further studies are needed to test the clinical utility of this approach in prospective patients."
In addition, the researchers noted that their study relied on ex vivo fine-needle aspiration biopsies, which don't fully recapitulate the standard use of fine-needle aspiration biopsies, and that it was conducted at a single center. They added that it needs to be validated at additional centers and in additional countries.