NEW YORK (GenomeWeb) – Researchers from the University of Cambridge have published new data that they hope will lead to the development of a clinical test that utilizes four blood-borne microRNAs to detect the presence of germ cell cancers.
The group shared results of a small initial study of samples from 25 children in the British Medical Journal earlier this month, in which they showed that a test that quantitates levels of four miRNAs can diagnose these tumors in different sites in the body, and can accurately distinguish malignant from benign occurrences.
Based on the study results, which the team hopes to replicate in a larger cohort, it appears the miRNA panel could also be used to assess the efficacy of treatment and monitor for recurrence of these tumors.
Germ cell cancers, as their name implies, arise from germ cells, which are the precursors of sperm and eggs. Mostly these cancers occur in the testes and ovaries, but can also emerge in other areas of the body like the brain. Current tests to detect these cancers and to distinguish malignant tumors from benign growths or other types of cancer are limited, University of Cambridge pathology professor and senior author Nick Coleman told GenomeWeb this week.
While there are some protein biomarkers that signal the presence of a germ cell tumor, they do not apply to all the subtypes equally. According to Coleman, finding a way to link the different types of germ cell tumors (GCTs) has been a longstanding goal in the field.
"These tumors are very heterogeneous clinically and pathologically but we always thought that there had to be some joint common biology to explain that they are all derived from the same cell of origin. We were interested in microRNAs as a marker of that shared biology," he said. "We do have protein-based markers, and they are pretty sensitivity when tumors do express the markers. But some tumors just don't."
In earlier work, Coleman and his Cambridge colleagues identified miRNA signatures in germ cell cancer tissue that not only appeared to apply to malignancies across the spectrum of these tumors, but also offered a method for highly sensitive diagnosis. After this initial discovery, the group then set out to see if they could measure the same miRNAs in blood or other body fluids for non-invasive diagnosis and disease monitoring, publishing an initial proof of principle for this in a single patient.
Other studies have also shown that serum miRNAs are sensitive and specific predictors of testicular GCTs in adults, but none had assessed them in a larger group of more diverse childhood tumors.
In their recently published study, Coleman and his team used a qRT-PCR pipeline to measure levels of a set of miRNAs from two clusters they had previously identified as being overexpressed in germ cell cancer tissue. They used 45 blood and cerebral spinal fluid samples from 25 children with brain or gonadal malignant germ cell cancers, controls with either non-malignant GCTs or non-CGT cancers, and several healthy controls.
The group identified four miRNAs from these two clusters that had the highest sensitivity and specificity in diagnosing malignant GCTs either from other cancers or from the absence of cancer.
"The [areas under the receiver operating curves] were in the 80 and 90 percents, which is very promising and is what we predicted from the tissue data that showed that levels of these microRNAs were very high in tumor, and very low normally," Coleman said.
Interestingly, the four molecules the group found to be most predictive in their pediatric cohort were also the same as those identified as predictive in other studies of adult GCTs.
Looking more closely at a single testicular GCT case, the researchers also showed that levels of some of the miRNAs in the two clusters could be used to track the disease's response to treatment and to monitor recurrence.
Analyzing 18 total samples collected from diagnosis through relapse, the researchers were able to see that the levels of several miRNAs dropped after the patient underwent treatment, and then rose robustly in parallel with a later relapse. Once treatment of this relapse began, the miRNAs then returned to baseline, the authors reported.
Even more interestingly, when radiological imaging later indicated the possibility of a second recurrence, the levels of these miRNAs remained low, the authors wrote. Examination of the tissue in question showed that it was a benign growth, not a malignant relapse, vindicating the miRNA-based finding.
According to Coleman, another equally important part of the study was the development of a robust and replicable pipeline for miRNA quantification to guide future development of a clinical assay. For example, Coleman and his colleagues wrote that based on their tests, the best way to most accurately detect and quantify these miRNAs in the context of GCTs is to perform two normalization steps using both an exogenous non-human spike-in cel–miR–39–3p and also an endogenous housekeeper miR–30b–5p, in order to obtain robust qRT–PCR quantification.
"We spent a lot of time trying to work out the best technical aspects — in terms of controlling for recovery of miRNAs in serum, for overall levels of circulating miRNAs that vary between people, and for if there is hemolysis — to quantify results in the context of all those [factors]," Coleman said.
Sharing these optimized protocols with the field will hopefully help maintain high standards and consistency as other research groups investigate miRNA-based GCT testing, and as the Cambridge group works towards further validating the sensitivity and specificity of their four-miRNA panel in a larger trial currently being planned, Coleman said.
"We feel that if this test is rolled out into clinical practice — which we are hopeful it will, and excited it could be — it will be really important to get the technical aspects of the assay completely understood and agreed," he explained.
Moving forward, Coleman and his co-investigators are now planning a larger study, hopefully with hundreds rather than tens of patients. This will allow them to establish sensitivity and specificity values for their miRNA panel with confidence intervals, and be able to truly evaluate the approach's clinical utility.
Additionally, Coleman said, the Cambridge team is in contact with other research groups who are looking at different types of blood-based biomarkers for GCT, namely circulating tumor DNA. The group hopes to be able to experiment in the future to see if ctDNA or other factors could add to the miRNA strategy they have developed or have other independent clinical value.