Wellness and Illness From a Precision Quantification Perspective
Blog Jul 29, 2019
A sliding scale exists in human health, where an individual moves between being clinically "well" to a "state of illness", and often back again.
When we are unwell, we typically present with symptoms, visit a healthcare professional and are prescribed a treatment that returns us to a state of "wellness".
How would our current approaches to healthcare management change if we could predict that illness was looming before it strikes, by precisely identifying where we sit on the scale of wellness and illness at any given time?
In this interview, Aaron Hudson, Vice President & General Manager, Global Marketing & Strategy at SCIEX, discusses why precision quantification, specifically quantitative mass spectrometry (MS), is the best approach for measuring the transition from "wellness" to illness.
Molly Campbell (MC): Why is it necessary to look to change our current approaches to managing illness?
Aaron Hudson (AH): We are all genetically different. Whether that be different races, different sexes, regardless: we respond differently to drugs and illnesses. This means that we need to tailor medicine. If you buy clothes from a store and they don't fit well, then you get them made for you. If we are serious about helping people and improving the human condition and human lives, we need to also tailor healthcare.
MC: How do you define wellness, and why is it important to measure it?
AH: I don't know whether I can define wellness personally - I think society has different views on that. Typically, if someone eats a nutritious diet and exercises, then they generally feel well, and their mental outlook on life is better. We are starting to see an explosion in the use of devices that measure our exercise output and even our duration of sleep, but that's not necessarily wellness. The subjective nature of wellness, and the fact that there are varying stages of wellness, means it is difficult to measure with current methods.
MC: What are the benefits of looking at health and wellness from a precision quantification perspective? Are there any drawbacks?
AH: You can't manage what you can't measure. If you can't measure your "baseline" wellness, then you cannot determine a shift from that baseline into illness, or measure your change in your baseline wellness as you get older. Unless you have a main metric that is a key performance indicator, a KPI in the business world, and you can measure that monthly, then it is difficult. We already look at certain biomarkers for health and wellness, blood pressure and cholesterol to name a few, but there just aren't enough of them, and we don't monitor them often enough.
If you think about the way that your body functions and break it down to the biochemical level, then you only really have genetic material, proteins, metabolites and then lipids. Essentially our body is made up of just those molecules.
Your genetic material does not change dramatically throughout your whole life, and certainly not on a monthly time frame that is measurable, and so if you really want to know how well you are then you need to look at proteins, lipids and metabolites that are fluctuating on a daily and hourly basis. The point is to try and get down to the actual molecules that are affecting your wellness, removing the background biological noise, and being able to measure those accurately and longitudinally. Quantitative MS is the best technology to do that right now.
MC: Please can you explain what is meant by precision quantification?
AH: There are two types of quantification: Absolute and Relative.
To give you an analogy: if you take a ring from your finger, there's an amount of gold in there. How do you find out how much gold there is? You weigh it. You take it to a jeweller, and they place it on a balance and they tell you that there is "X" amount there. But what if the balance is wrong? We can be confident that the balance is not wrong because it should have been calibrated and traced back to a “gold standard” which is actually defined by a kilogram ingot called “Le Grande K” that is locked away in a safe in Paris. There is a process of traceability and accuracy in that measurement, and this is absolute quantification.
The other side to quantification is relative quantification, where you compare two or more samples with one another. That tends to be the type of measurement used for research purposes to uncover new disease biomarkers. For example, taking a cohort of disease samples and control samples and observing which proteins, lipids or metabolites are up or down regulated in these samples. This gives us a window into what molecules are changing in the disease compared to the wellness state and will help us to discover new wellness and disease markers in the future.
MC: What technologies are required to measure both wellness and illness in a laboratory?
AH: It all starts with MS. It is exquisitely sensitive and specific which is highly important as you have to quantify the right molecular isoform, free from interferences.
When you think about accuracy, MS is the only technology that measures the molecule. Techniques that use antibodies such as ELISA and Immunoassay don't actually measure the molecule, they measure the antibody as a proxy for the molecule, and as such can be prone to interference and inaccuracy. SCIEX has spent the last 50 years trying to optimize MS to have the highest sensitivity and specificity for the quantification of molecules.
Depending on what the focus of the lab is, there is a continuum that starts with the research into new biomarker discovery, moves through biomarker validation, and then into clinical utility.
For biomarker discovery, we have a technology called SWATH on our TripleTOF mass spec systems which can quantify 5,000 proteins from a cell line in one hour. So now, you have a research instrument that can quantify thousands of molecules in hundreds of samples. Researchers around the world are using this to understand wellness and illness at the molecular level. If you are running large numbers of samples in your study, you need sophisticated software to deal with that volume of data. The OneOmics software is cloud based and can compare hundreds of samples relative to each other, and even start to integrate and draw information together from metabolomics, proteomics, proteogenomics, etc., so called multi-omics analysis.
After this discovery stage, you theoretically know what the biomarkers you need to quantify are, allowing you to use a targeted analysis that offers better precision and accuracy. Then for both validation and clinical utility, you typically would use a Triple Quadrupole 5500+ or 6500+ LC-MS/MS System to get absolute quantification of over a hundred analytes in a single analysis that lasts a few minutes. Our QTRAP technology can quantify 1,100 lipids and, when combined with microflow, over 300 metabolites within a plasma sample.
So, a laboratory needs the hardware, triple quadrupoles, and some form of liquid chromatography upfront. It also needs software with sophisticated algorithms for peak detection and quantification. It is important to be able to integrate the area under the peak accurately and reproducibly with minimum manual interrogation. This is something we have been working on for more than 20 years.
MC: Why is precision quantification a key factor of consideration when developing novel technologies, and how exactly do we achieve high levels of precision?
AH: Precision and accuracy are two different things. You can be precise and not accurate. Precision means how close two or more measurements are to each other, whereas accuracy means how close you are to the standard or known value. Imagine a dart board – you're aiming for the bullseye. If you are precise and accurate then all the darts end up in the bullseye. If you are only precise then all the darts end up clustered somewhere else on the board. If you are only accurate then the darts are spread around the bullseye but are not clustered together. In our quantification with mass spec we need to be both precise and accurate, but the other question is “why is quantification important”?
Well, if I told you that you had won the lottery, your first question would be "how much"? Have I won a lot or a little, and “a lot” can mean different things to different people. The more accurate and precise you are with the amount you have won, then the better decisions you can make.
A more health-related example of that is if you were to have an organ transplant then you receive immunosuppressant drugs for the rest of your life. These drugs suppress the immune system to ensure you do not reject the organ you receive. However, if you are given too much of the immunosuppressant drugs, they are toxic, and they will kill you. If you are not given enough of them, you will reject the transplant, and die. There is a narrow window of tolerance, and if you can’t quantify the amount of drug in your body precisely and accurately then the repercussions are significant.
So now we can quantify molecules with precision and accuracy, we need to account for biological variability. We are all biologically different. If I measure one molecule in me and one in you, say the iron storage protein ferritin, we may not have the same levels of that protein. We still function, we are just genetically different. Biological variation and biological noise is pervasive throughout any type of scientific study and has been well documented in the scientific literature.
Lastly, there is analytical variability and the need to get the same result in any lab in the world, on any given day, for the same sample. This is not as simple as it seems, and scientific reproducibility has gained a lot of attention and focus recently. As instrument vendors, we need to do more in attaining tighter specifications in instrument to instrument variability that can help analytical scientists to address this.
One of our latest products at SCIEX is the Optiflow® Source that can be retrofitted to most of our mass specs, and this eliminates the "front end" human optimization that causes much of the instrument-to-instrument variability that is seen in labs – it’s essentially plug-and-play and gives you the same results that an expert user would.
MC: Are precision diagnostics, precision therapy and companion diagnostics the future of healthcare? If so, how are they integrated in an individual's journey from wellness to illness and in return to wellness?
AH: I see a future, that I think is shared by many in the translational medicine community, whereby when a baby is born it will have his or her genome sequenced. From this sequence you will predict that the baby has the propensity for say 5 diseases during his or her life. But the DNA indicates the susceptibility, not the onset of the disease, so for each of the diseases there will be a panel of protein, metabolite or lipid biomarkers that will be monitored to determine the onset. There will also be a set of biomarkers that are monitored during that baby’s life that are typical of wellness, and together these make a unique signature that will need to be measured monthly to establish a baseline.
Monthly monitoring using current practices is invasive and it's time consuming; people don’t want to go to the hospital every month to have an armful of blood drawn. With mass spec you can measure hundreds of analytes in microliters of sample, such as a pin prick of blood. A drop of blood is the type of sample that can be acquired in the comfort of your own home and sent in the post to a reference lab that can automate the testing based on your own personal wellness signature. This is where we are talking about precision diagnostics using multiplexed MS that is very specific and highly accurate.
If you repeat this monthly you will generate a personal longitudinal baseline that can smooth out any natural variation in the levels of analytes depending on life circumstances, diet choices, a common cold, etc. But at some point changes in those markers will be indicative that this person is moving from wellness to illness. The data from millions of people will also be anonymized and aggregated for machine learning to continuously improve these predictions.
Once you have the early signs of a shift from wellness to illness a physician will intervene with a precision therapy. There will be multiple biopharmaceutical options, but based on your genome there will be ones that will work better – we treat breast cancer this way now, based on a positive or negative score for the oestrogen, progesterone and HER2 receptors. These types of test to determine the treatment path fall under the umbrella of a companion diagnostic.
But, how do you know that this specific treatment is working? There are very few objective and systematic tests available to determine this other than long-term observation of the patient, but by quantitative measurement of molecular markers there will be a means to determine whether a patient is returning to their wellness baseline or not. Some markers may not return to the original level, and this may set a new baseline. This type of regular therapeutic effectiveness monitoring is different than a companion diagnostics and I predict will be an area of intense research in the future.
This is really what precision medicine is. It’s a broad term for precision diagnostics followed by precision therapies coupled with companion diagnostics and therapeutic efficiency monitoring. But you don’t get precision medicine, precision diagnostics or precision therapies without the precise and multiplexed quantification of molecules – most likely using MS.