Metabolomics for Sepsis: Predictive Biomarkers
Hi everyone, welcome to this podcast by Cambridge Healthtech Institute for the Molecular Diagnostics for Infectious Disease meeting, part of the Molecular Medicine TriConference, which runs March 7 – 9th, 2016 in San Francisco, California. I'm Kaitlin Searfoss, Associate Conference Producer.
We have with us today one of our chairs for the meeting, Dr. Raymond J. Langley, Assistant Professor, Department of Pharmacology, of the University of South Alabama. Dr. Langley, thank you for joining us.
Can you summarize the work you are currently doing on Metabolomic risk analysis models for Sepsis?
The study goes back about 8 years ago, when I was a Postdoc, working for my mentor, Steven Kingsmore and the idea was that we wanted to take a systems biology approach to understand the risk factors for patients that die from Sepsis or survive and some of the progression markers. We looked at a number of different Omic techniques, looking at Metabolomics, Transcriptomics, as well as Proteomics and surprisingly, what we found was that the Metabolomic analysis that we had performed at Metabolon which is in Carey, North Carolina ... What we found was that they were the most predictive in determining patients that would live or die.
Ultimately, what we did is we selected a few markers using a number of different logistic regression techniques, which were very predictive for outcome, and these markers, which were basically acyl carnitine esters, looked to be definitive of Mitochondrial dysfunction and kind of suggested to us that what we were seeing is patients that were dying were in an energy crisis and were unable to convert, basically, fatty acids into energy to survive this energy crisis.
We developed that model and published a paper in Science Translational Medicine in 2013. I continued doing work when I went to Lovelace Respiratory Research, looking at non-human primates and this is another study where we combined the Metabolomic analysis from the animal's plasma with changes that we saw in the lung transcriptome. In utilizing this technique, we were able to further refine the model. It pin-pointed that it looked like mitochondrial dysfunction and basically, a problem with beta oxidation, but we were also able to link a number of other, different markers, such taurine-conjugated bile acids. We also looked at acyl GPCs which are basically long chain fatty acids on a glycerol body with a phospho-head. These were also predictive of Sepsis and then the final markers were basically metabolites from the urinary pathway.
By doing this, we were able to further refine the model, and really get very strong prediction of survival in patients that had Sepsis. This will, again, prospectively analyze a number of patients that are enrolled at the emergency department and will follow them through, whether they go to the ICU or to the floor, and will try to determine how well our model is at predicting survival versus death, as well as seeing how it links to other organ dysfunction and are they predictive of organ dysfunctions.
What would you identify as the biggest challenges facing scientists, in terms of developing technology for metabolics for Sepsis diagnostics?
That's a good question. There's a number of different problems. Initially, right now, I think the biggest problem is trying to get the platforms to analyze these analytes as quickly as possible in a very short period of time, especially for something that we're looking at, which is a very acute disease and treatment. They call them the golden hours. You have about 8 hours from the time a patient develops septic shock to get them normalized, or the patient is likely going to have a very bad outcome. That's the biggest challenge for us, right now. It's trying to get a platform that we can get these analysis done quickly.
Another challenge, I think, that many researchers will have, and this is for the lack of understanding of the basic biochemistry. This took me about a year to really catch up and understand what all of the different biochemical pathways are.
The final thing that I think is really difficult is, even though you may understand the biochemical pathway, you still have to link it to the organ dysfunction, so which organs are actually producing the metabolites that are changing. What we found is by doing integrative Omics or marrying the metabolomic signatures with Transcriptomic or Proteomic signatures, it helps you to identify the biochemical pathway, as well as, get a better understanding of which organs or tissues are being affected and are actually producing the metabolites that you're looking at.
Who are you most looking forward to hearing from, at the Molecular Diagnostics for Infectious Disease Conference?
I think a little bit of everything. One thing, I'd like to talk to some of the people who are developing any of the new tools for analyzing metabolomics. Are there any new platforms that are coming forward that'll be able to do it quicker, maybe simpler, for technicians to run these studies? I also would be interested in talking with a lot of the clinicians who may be interested in utilizing this and what their thoughts are on how effective these markers would be, compared to what they're currently using. Right now, when a patient presents, lactate is one of the big markers for outcome. It's a very good marker. When you start to see it going up, you know that the patient is probably in Hypoxia and if you don't normalize it, the patient's going to have a very poor chance of survival. This is currently the gold standard. We believe that our markers could be much better, so I'd to basically present some of that data and get some feedback from the clinicians and whether they think this is a diagnostic that would be useful for them.
Thank you for your time and insights today.
Great, I appreciate it.
That was Dr. Raymond J. Langley, Assistant Professor, Department of Pharmacology, at the University of South Alabama. He'll be speaking at the Molecular Diagnostics for Infectious Disease event at the upcoming Molecular Medicine TriConference, taking place March 7 – 9th, 2016 in San Francisco, California. If you'd like to hear him in person, go to triconference.com for registration information and enter the key code 'podcast.' I'm Kaitlin Searfoss. Thank you for listening.