Jennifer Van Eyk has consistently made the case for proteomics as an essential foundation for precision medicine. From scalable LC-MS workflows to single-cell analysis to the biological heterogeneity that shapes disease and drug response. 

Here, we ask what it will take for precision healthcare to become a clinical reality: the limits of genomics-first thinking, the financial and regulatory barriers to biomarker adoption, and the role of remote sampling, preventative care, and more personalized drug development in the future of medicine.

What is the current state of precision or personalized medicine? 

As a participant in several national healthcare initiatives, it strikes me that much of the conversation is focused on genomics and genetics. These are of course crucial for a lot of things, but they certainly don't cover everything in biology, especially not therapy response or the myriad of environmental and life factors that also play a huge role.

A real turning point was when we published a paper in NatureCancer on cancer prediction after diagnosis. We were studying pancreatic cancer, and I was genuinely surprised and delighted that proteins, both in plasma and tissue, were overwhelmingly driving our modeling results to predict disease survival. Genetics, on the other hand, played almost no role. Now, this might not hold for every type of cancer, but the dominance of proteins was unexpected and exciting – especially considering we were looking at basic protein data, not even diving into post-translational modifications or other complex markers. My expectation is that as we start looking into more detailed proteomic profiles, such as protein post-translational modifications, we’ll see even more relevance. Proteomics is poised to drive precision medicine forward, perhaps even more than genomics.

There’s such a breadth of biologically important data that proteomics can provide, from conformational biomarkers to newly synthesized proteins, protein complexes, and even in vivo interactions. It’s surprising how much remains overlooked in these big projects and consortiums that are generating and analyzing data.

That said, it’s still tough to get biomarkers into the clinic. And I don’t think it’s a technology issue; it’s more about financial models and regulatory hurdles. These are complex problems, largely unrelated to technology, that need to be addressed if we want to see these biomarkers routinely used in clinical settings. Furthermore, when we talk about personalized medicine, people often think about diagnostics for early-stage disease or health monitoring over time, which can be ideal. But for this to be successful, central questions remain: who pays for it, how is it implemented and reported to patients, and what is actionable? 

In our lab, we’re exploring disease progression and natural disease history using remote blood sampling devices to expand the biomarker domain. However, realistically, large scale broad adoption of precision medicine applications are still a way off. Where I see it taking off first is in diagnostic work: when a patient is already sick and has a diagnosis, biomarkers could be used to select the best therapy. In other words, individualizing treatment after diagnosis, as opposed to tracking disease progression or health status over time. Unfortunately, tracking personal health over the long term remains challenging for most individuals for multiple reasons including clinical, financial and practical barriers.  

Like others, we are trying to incorporate personalized diagnostics and individualized therapeutics. Hence, we are developing scalable proteomic pipelines for therapeutic screening –looking for new drugs and targeted treatments more precisely. For instance, if a current drug works for 80 percent of patients, we need to understand the mechanisms that would work for the remaining 20 percent, and maybe even develop drugs tailored to those mechanisms. Thus, personalized medicine means having the diagnostic data to pinpoint the disease mechanism(s) and repurposing or developing the drug to match that.

How can we make drug development more precise and personalized? 

It’s easy to see how, with drugs currently used in cancer, there’s often a small subset of patients who respond very well, while others don’t at all. Identifying those subsets is essential for a personalized approach to work. But this can also apply outside of cancer, for example with prescription of statins. People are commonly prescribed statins even when they are fairly low-risk of having a cardiac event, and in this case, statins will have a life-saving impact for only a very small percentage of patients. There is also controversy over potential side effects, which must be balanced with the varying potency of different medications. After all these years, we are still refining our understanding of who should be on these drugs.

There are two main aspects to consider. One is where we focus on determining who will genuinely benefit from a specific drug. Second, is the emerging concept that for many conditions, such as cancer, brain injury, even heart disease, there could be multiple underlying mechanisms driving the same clinical phenotype. This is becoming increasingly evident across a wide range of conditions, and it will become important to identify the molecular drivers of a person’s disease to prescribe the appropriate therapy. 

In our current work, we’re exploring single-cell proteomics, and what we’re seeing is fascinating. Even among cells that are supposedly the same cell-type, say cardiomyocytes, not every cell responds the same way to a given drug. This could explain some of the variability in therapeutic responses. For instance, a drug might be working well for a patient overall, but perhaps it’s only effective for half of the target cells, while the other half are in a non-responsive state. This means that when a drug is not effective for a person, we do not know if it is the drug per say or the cells not being in a responsive state.

What are the main barriers to making precision medicine a reality? 

A big challenge for mass spectrometry-based proteomics, such as DIA (data-independent acquisition) on instruments like an Orbitrap Astral, even with spiked-in standards, is that we cannot yet reach the precision needed for clinical use (under 5 percent to be viable clinically). Large, targeted assays, however, meet those standards, but they require specialized skill sets and optimization for every peptide, which can take a long time. Although discovery is becoming standardized, targeted assay development is far more complex and involves a different type of training. As of today, our field of proteomics does not have enough trained targeted assay specialists. Unfortunately, the skillsets required to go from discovery to clinical application are not typically housed in the same labs or even in the same institutions, which hinders the transfer of candidate biomarkers up the translational pipeline.

For precision medicine to fully enter clinical practice, we need to train more specialists in targeted quantitation and at the same time, ensure these assays become more standardized, accessible, and stable over time. There has been great progress, but we’re still not fully where we need to be. This divide between discovery scientists – those focused on targeted quantification – and the clinicians using this data is challenging. Bridging these areas will be essential to make precision medicine more patient-focused.

What might precision medicine look like in 20-30 years? What’s your dream scenario?  

My big dream is that we will incorporate remote blood or plasma sampling devices into research and clinical areas. This will lower the barriers to people accessing their own biochemistry – whether to monitor their health over time, identify when immediate treatment is needed, or determine when a change in treatment may be necessary. Imagine you’re undergoing cancer treatment and live in a remote area, or you’re a single mom with limited time; instead of making a long trip to the hospital, you could use remote blood/plasma sampling to monitor your health. This approach could play a transformative role globally – not just in the US – by providing better access to healthcare and empowering people to make good clinical decisions.

Remote blood sampling can also reduce the burden of gathering data for clinicians, and I would love to see a future where, instead of going to a clinical diagnostic lab or making a physician visit, you could get a blood sample or other test done from home. This would allow for more frequent testing at any age, ensuring samples are analyzed consistently. Virtual interactions with physicians could also increase, providing more continuous and meaningful patient-physician interactions, which I believe is beneficial. In addition, more frequent data collection could help reduce costs around missed diagnoses or unnecessary treatments.

On a broader scale, people should have the ability to own and understand their own health data. Access to good medical care is still challenging for many, and there’s nothing worse than not having a diagnosis or not knowing if your treatment is being effective. Remote blood/plasma sampling could provide timely data to support clinical decisions and make good healthcare more accessible. That’s my dream: for these tools to be deployed as a seamless part of the virtual world of medicine, allowing people to make informed, real-time decisions about things like drug adjustments or changes in therapy.

While this doesn’t necessarily change diagnostics itself, it does ensure that healthcare is personalized and up-to-date for each person, regardless of their socioeconomic situation, location or environment. That’s a fundamental part of precision medicine to me: the equal access to data-driven, personalized care by reducing barriers.

What role could preventative care play in the future of precision medicine? What about nutrition? 

Food provides essential nutrients that our body needs, and it plays a crucial role in health. Integrating nutrition into a healthcare program could absolutely be beneficial, and it could even be done virtually, which would help reduce costs. Imagine talking to your nutritionist or other health professionals remotely; I think that would be a much better approach.

But, as always, the question is who pays for this kind of preventative care. Right now, if someone is already sick, the payment schemes are in place, but when it’s about keeping someone healthy, it becomes a more complicated question. Ideally, I’d love to see this kind of care from birth onwards, but practically, this continual oversight may not be achievable, and interaction may start only when a person become under medical care for a specific reason. Say someone is about to have knee surgery, has just been diagnosed with cancer, hypothyroidism or adrenal insufficiency, then they could be monitored over time. That could have incredible value. Another example would be to use remote blood/plasma sampling after a major event, like a heart attack or heart transplantation, which could be hugely beneficial. Yet, even with technologies and advocacy in place, identifying the payer will remain a challenge for these scenarios. 

Jennifer Van Eyk is Professor, Cardiology and Basic Medical Sciences, Director of Advanced Clinical BioSystems Research Institute. Cedars-Sinai Medical Center, USA

Newsletters

Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.

I have read and understand the Privacy Notice*

Subscribe