Clinical researchers at Beth Israel Deaconess Medical Center and Harvard Medical School have teamed up with RainDance Technologies to explore the use of RainDance's picodroplet-based digital PCR platform to measure microRNA biomarkers in the blood of patients undergoing various cardiac therapies.
Specifically, the researchers recently finished one study that uncovered miRNA biomarkers that may eventually help physicians determine how favorably patients with advanced heart failure respond to cardiac resynchronization therapy, Saumya Das, an assistant professor of medicine at Harvard and BID, told PCR Insider this week.
The team is now working with RainDance to see whether the company's RainDrop Digital PCR system has the required sensitivity and multiplexing capability to serve as a potential clinical assay modality to measure these miRNA biomarkers.
In addition, Das is principal investigator on a follow-up project that last month was awarded $500,000 from the National Center for Advancing Translational Sciences for the first year of a five-year project eligible for up to $4 million in total NCATS funding.
In this project, Das said, the group is using RNA sequencing to identify potential miRNA biomarkers of adverse mechanical and electrical remodeling in patients following myocardial infarction. The group hopes to also eventually use digital PCR to develop an assay for these biomarkers in MI patients to help guide treatment, Das said.
Das said that the related projects got their start a few years back when his group began investigating whether it could identify baseline levels of miRNAs that could predict which patients with advanced heart failure would respond favorably to cardiac resynchronization therapy.
"If we could actually figure out who was going to [respond] better, then you could tailor the therapy toward those patients and eventually this would be cost-effective as well as save patients from procedures that don't benefit them," Das said.
When they began their project, RNA sequencing had not yet been established as a reliable research tool, so the group used TaqMan Array Human MicroRNA Cards from Life Technologies to identify a group of differentiated miRNAs in a very small set of patients that had either responded favorably or unfavorably to CRT.
Then, using conventional real-time PCR, Das and colleagues validated those miRNA biomarkers in 40 additional patients, and were able to whittle the biomarkers down to just "one particular microRNA that was, independent of all clinical variables, a pretty good predictor of how patients would respond in about six months … after the CRT," he said.
Since then, the researchers have looked more closely at the biology of this particular miRNA and are convinced that it does appear to play a role in patient response.
Then the group began thinking about how it might translate this finding into clinical use, which is how it began working with RainDance to see if its picodroplet-based digital PCR technology had the necessary dynamic range to detect the microRNAs in plasma samples.
"Obviously the levels of some of these are very low, so we were interested in figuring out how sensitive digital PCR would be, and how well it correlates with traditional PCR," Das said. "And the third thing we are interested in is whether it can be multiplexed, because as we moved along we noticed that there were other microRNAs that changed … coordinately with the one we were interested in, and we were hopeful that we would find other microRNAs … that might be even more powerful as a multi-marker type approach to predicting things."
Indeed, researchers from RainDance and the Université de Strasbourg in France have previously described a modified version of RainDance's digital PCR technology that could multiplex based on fluorescence color and intensity. They also used the method to develop a five-plex assay for spinal muscular atrophy using just two fluorophores (PCR Insider, 5/26/2011).
According to Das, his team's collaboration with RainDance is ongoing in this area, and preliminary results "look very good," he said. "We get good agreement. We can take a synthetic microRNA and figure out exactly how many molecules we are putting in, and we … are able to accurately quantify … when we do digital PCR," Das said. "And one of the [RainDance] scientists that we are working with has been able to multiplex two or three different miRNAs, and gotten pretty good results, too."
Das said that this work will hopefully be the subject of two forthcoming publications — one describing the microRNAs and their biology, and the other describing the molecular biology techniques that were used for discovery and validation.
In the meantime, under the new NCATS grant, the researchers will try to answer an even broader question about whether miRNA biomarkers can be used to predict patients who will suffer adverse mechanical or electrical remodeling, which can lead to heart failure and sudden cardiac arrest in patients who have previously experienced an MI — even if their heart function as revealed by current clinical tools appears to be sound.
"The current tools available to risk stratify patients for both adverse mechanical and electrical remodeling are fairly crude," Das said. "One of the things people use is a crude echocardiographic assessment. That has certain drawbacks; for example, the current Medicare criteria for putting a defibrillator in patients who've had MI, is anyone whose heart function is less than 30 percent … qualifies to get a defibrillator. But if you actually look at those patients and see how many of them actually use the defibrillator, it's less than a third. Conversely, many patients … don't have such bad heart function, but nonetheless have lethal arrhythmias. And there is no way to target those patients, because we don't have the tools for that yet."
In this study, the group is working with Brigham and Women's Hospital to compare a small group of similar well-phenotyped patients who have experienced MI and see if it can identify miRNA biomarkers in the blood using Illumina RNA-seq.
"From this we hope to get a bunch of candidate miRNAs that are different between these two cohorts, and then validate this in a larger 374-patient cohort to figure out whether we can get miRNAs that are sort of surrogate markers of this electrical and mechanical remodeling," Das said.
Then, he said, the plan is to develop assays for those candidate biomarkers on the digital PCR platform, "and then look at outcomes in a 4,000-patient cohort. In the end we think this will be something like 10 to 12 miRNAs."
Das said that the group believes droplet-based digital PCR is the best approach for this assay development, as the cost of sequencing has not yet dropped enough to make it feasible, and real-time PCR — while often time sensitive enough — has additional drawbacks, particularly when it comes to absolute quantification.
"There are levels [of these miRNAs] in normal patients, as well, so I think that quantification becomes important," Das said. "We also have a normal cohort, so once we find out what these microRNAs are, we want to look at the normal cohort and see what the distribution is. And I think there, having absolute numbers per milliliter of blood would make a biomarker test more sensitive. Because this is not like cancer, where there is a mutation, and either it's one or none – here it's probably a range in normal patients and a range in patients who do poorly … and who do well. So that will be one of the strengths of digital PCR."