When will we get to a place of reliable protein biomarkers as predictors or indicators of disease? Ole Vorm is a founder and director at Evosep, producing separation products between the patient sample and a mass spectrometer.
What is the state of those analyses and what are the obstacles yet to be overcome?
The first obstacle in terms of the proteome itself is that in any sample - urine, blood etc.- there are a few abundant proteins that represent the bulk of the material. It’s likely that the proteins of interest (if we can find them) are a small portion of the total and any changes in quantity or relevant modifications may be subtle.
But of the currently FDA approved biomarkers typically analyzed by ELISA assays, about 50 can be detected by simple injection of a protein digest into a mass spec. That’s good but not a lot.
From a physician’s perspective in the clinic,
…they never start out, uh, from a blank sheet of paper. They have an idea what they're looking for and asking for quantitation of say, two or three biomarkers, I mean, proteins basically,would typically suffice to say, okay, we're in this direction, versus going in that direction. And that is very cheaply, automatically done with ELISAs running on a fully automated, robotic platform in the clinical biochemistry labs.
The current ability to detect 300 proteins out of the mix is way more than a physician could use right now, but probably does not go deep enough to uniquely detect a disease. In many sports, this is called no man’s land.
Nevertheless between improvements in detection, robustness and AI to help analyze the data, Ole sees the field moving dramatically over the next few years.
In terms of analysis, we may find that the changes we see in a patient’s proteome, may not directly identify targets for therapy. Rather they might reflect secondary effects of a disease. In this case, where treatments can be found, those biomarkers may serve to monitor progress during treatment as opposed to being used for diagnosis.
Beyond the need for improved robustness (a clinical analyzer needs to run hundreds of samples per day without human intervention), there aren’t enough mass spectrometers to analyze the possible volume of samples. Using proteomics on top of genomic or metabolic analysis seems to be a more likely strategy.
Ole closed with this idea:
I do think things are progressing and, and I think, you know, what is really, really needed in our field would be sort of like the first success story. I mean that, from my perspective, that's what we really need to somebody, some group coming out and saying, so now we have this, this protein pattern is the detection of this and this. Whatever the success story basically is, I think it is just important that we begin to see some success stories and that will fuel the fire and get, more things going.
I’m interested in your thoughts. Do you know of any success stories already? Where do you see proteomics taking off?
Chat with Chris about custom content for your life science brand.
Share this post