Predicting Mortality in the ICU

A Q&A with Angela Rogers, MD, MPH, Brigham and Women's Hospital, Channing Laboratory, Boston, Mass.

Section Sponsored by:

A Q&A with Angela Rogers, MD, MPH, Brigham and Women's Hospital, Channing Laboratory, Boston, Mass.

ADVANCE: What are metabolic biomarkers?
Rogers: I come from a background of doing genetics. When you think about genetics, you think about diseases like cystic fibrosis. One gene can affect a person's health quite a bit, CS is a great example. In the case of more complex traits, like asthma, or things you see in the intensive care unit, such as sepsis, it's much more complicated. We're quite sure that it's a matter of hundreds of genes interacting with each other. That's something I'd say we've learned from the human genome in general. I'm sure it's the same thing for the diseases we study in the intensive care unit, such as respiratory distress syndrome, or sepsis are very complex. You have genes that lead to gene expression, which lead to proteins. As those proteins are working, they give off products called metabolites. It's the last step in the process from genotype to phenotype, you have proteins and metabolites that are hard to study, but interesting.

Their genetic code has been with them since childhood. A powerful reason to study proteomics or metabolomics is that it gives you a snapshot of what's going on at that moment. It might be a great marker for how you can tell if someone is really sick and needs the intensive care unit or who needs to be in a clinical trial, or if someone is doing OK, and might be fine, who might not require as much attention.

ADVANCE: How did you make the connection between metabolic biomarkers and ICU mortality?
Rogers: The best marker that we have as a signal that someone's really sick is a metabolite called lactate. That's our best ICU biomarker to date. Lactate is a breakdown product of the anaerobic cycle, which tells you the body isn't getting enough oxygen in certain parts.

Lactate is something that is very frequently checked in emergency departments throughout the country. I'd say that's one of the best things we've got as a biomarker. It hadn't been studied a lot in the past, both in proteomics and metabolomics because it was very hard to do it quickly and accurately. Many strides have been made from this technology that now allow those things to be measured really well and to get a full picture of the metabolism.

ADVANCE: What markers are you focusing on?
Rogers: Our project was on a discovery panel. Meaning, we don't know why lactate is important, but we don't what else is. We looked at lipid breakdown processes, amino acids, multiple types of metabolites -- peptides, lipids, carbohydrate breakdown products. It's not possible to study all metabolites; we don't even know what they all are, they are still being discovered. We looked across a lot of classes to see what might be.

ADVANCE: What will this mean for patient care?
Rogers: I would say that we, and other groups, we were able to do this in two groups. Our cohort of patients in our ICU, and then another group of patients in three hospitals across the U.S. that had serious infections in the ED, and were able to show that the same metabolites seemed to show who was going to die in our ED population and their population. I think it really gives you a sense of the promise of metabolomics. It's clear that this is an area that has promise and needs further research. If we could get these results back right away in the ED, when someone comes, maybe give an early warning signal. A way to say, "Hey that patient is sicker than you thought."

Doctors and nurses are very good at recognizing people who are sick based on vital signs, etc., but it's not perfect. Sometimes you'll see someone who looks really good at the present time, and isn't admitted to the ICU, but then gets much sicker later. The body is very good at maintaining itself so you have warning signals about the heart rate or blood pressure falling, but the body is doing its best to sustain itself in a dire situation. This would be another marker of who's the sickest. That would help with triaging patients by making sure everyone was on their guard. This person's metabolome is really distressed, but the other vitals don't look that bad. The vital signs are OK, but the patient might be sicker than we thought.

This has also made it easier to identify subsets of patients are who are good candidates for clinical trials. Many studies have shown that ICU mortality rates are falling, which is a great thing. That makes it harder to find the sickest patients to enroll in clinical trials. The ones who are getting better based on fluids shouldn't be in clinical trials for new drugs. Most people who come to the Intensive Care Unit are very sick, but the vast majority are going to get better, which is a great thing. Can we figure out who those patients are that won't get better, despite our best efforts?

ADVANCE: How do you do this testing? 
Rogers: Ideally, a blood panel is run once a patient is admitted. There was a study done with patients who agreed to give us blood samples just so that we could learn about them. Later on, we could send it to a company who would check the metabolomics would learn about this signature for who was the sickest, and who was not. This was a research study. The next step is between our collaborators is start thinking about technologic advances. We did this study at the Bringham, then we collaborated with people who had samples from Michigan, and North Carolina. We all used the same company because they had the same metobollic profiling. Lactate testing goes on in ICUs every day, metabolomic testing does not yet.

ADVANCE: Why is this important for laboratory testing?
Rogers: We are looking for better ways to take care of our sickest patients. Patients who are our highest risk for complications, dying from sepsis, or having lung inflammation, or other complications. This is an area that has been understudied. It was hard to do this testing before because the technology didn't exist, plus the testing was very expensive. But, now it is possible because of advances in technology, and has a lot of promise to help us with classifying patients a little bit better. 

ADVANCE: What does this mean for the future of laboratory testing?
Rogers: With the fall of cost in this testing, it makes it possible and more accessible. If we can really prove that this matters, that identifying these things early on can help you with finding at risk patients, then this could save lives. It's now feasible that a lab could do this in the next few years, theoretically. For the future of this testing, figuring out what really are the best markers for the various diseases would make a huge impact.

Scott Hunt is a former ADVANCE intern.

Articles Archives


Email: *

Email, first name, comment and security code are required fields; all other fields are optional. With the exception of email, any information you provide will be displayed with your comment.

First * Last
Title Field Facility
City State

Comments: *
To prevent comment spam, please type the code you see below into the code field before submitting your comment. If you cannot read the numbers in the below image, reload the page to generate a new one.

Enter the security code below: *

Fields marked with an * are required.


Back to Top

© 2016 Merion Matters

2900 Horizon Drive, King of Prussia PA 19406