Precision medicine by finger prick

Precision medicine is often described in terms of AI-assisted diagnostics, molecular dashboards and treatments tailored to the biology of a single patient. But one of its most practical breakthroughs may arrive in a far humbler form – a finger prick, a paper card and a stamped envelope.

That is one of the most striking lessons from a new preprint from researchers in Stockholm, who used repeated self-sampled dried blood spots from 808 young adults to track how the immune system shifts over time. The study, Molecular profiling of repeated self-sampled blood reveals dynamic immune phenotypes in young adults (preprint under review), followed participants in the BAMSE cohort during the pandemic and combined serology, autoantibody measurements and proteomic profiling to capture how infection, vaccination, genetics and physiology shape the biology of blood over time. The broader message is that precision medicine may depend as much on when and how often we measure biology as on what we measure. 

Just as important, the study is also a case study in cross-institutional collaboration. This project brought together the long-running epidemiological power of the BAMSE cohort and Karolinska Institutet (KI), the clinical reality of Södersjukhuset and Sachs’ Children and Youth Hospital, and the molecular technology and proteomics expertise of KTH Royal Institute of Technology at SciLifeLab. It linked patients, population data, clinical questions and advanced measurement technologies into a single workflow. 

A unique cohort with a long memory

The foundation for all of this is BAMSE, one of Sweden’s most valuable longitudinal health resources. The cohort began in 1994 and has followed more than 4,000 children born in Stockholm between 1994 and 1996. Over the years, it has grown from an allergy and asthma study into a broader platform for understanding how genes, environment and lifestyle shape health across the life course. BAMSE is a collaboration between KI and Region Stockholm, and has already generated 51 doctoral theses and more than 350 scientific papers. That long memory is what gave this pandemic-era study unusual depth: the researchers were not studying anonymous samples, but young adults whose health history had been mapped since infancy. 

– We have a unique data set in the BAMSE cohort, says PI Erik Melén, MD, and a professor in paediatrics at the Department of Clinical Science and Education, Södersjukhuset, KI. 

– We followed 4,000 study participants. They were newborns when they were recruited, and now they’re in their 30s, Erik continues.

In this study, he explains, the long perspective became especially powerful because repeated dried blood spot sampling could be layered onto the cohort during and after the COVID-19 pandemic. The result was a time-resolved record of changing molecular states.

Why timing matters in precision medicine

Population studies of blood proteins have already revealed a great deal about human biology, but one major limitation has remained: transient biological changes are easy to miss. A standard clinical blood draw tells you what is true at one moment. It may miss what happened last week, what changes after vaccination, or what briefly flares after infection before settling back toward baseline. 

Jochen Schwenk, a KTH professor of translational proteomics, platform scientific director for SciLifeLab’s Proteomics Platform, and currently a visiting professor at Sachs’ Children and Youth Hospital, Södersjukhuset, has pioneered the study of circulating proteins for precision medicine, even from minimally invasive, self-sampled dried blood spots. In the interview, he describes why the BAMSE partnership mattered so much.

– Early on, we realised that if we work together as a team, we can get much further, Jochen says. 

– Everyone brought complementary expertise to the table, he continues.

He points to a common problem in molecular biomarker studies: “One of the missing aspects has often been that we don’t know about the health history of the donors.” 

A finger prick instead of a clinic visit

Rather than asking participants to repeatedly attend a clinic for venous blood draws, the researchers relied on self-administered blood samples on filter paper. Maura Kere, MD, who recently completed her PhD at Karolinska Institutet, and one of the study’s authors, puts the advantage plainly:

– You don’t need healthcare personnel to actually take the blood sample. You can ask individuals themselves to do a simple finger prick, which anyone can do at home, says Maura.

– What that helps with, she says, is repeated sampling over time: the ability to “capture health fluctuations”, including infection and other short-term changes. 

Annika Bendes, a KTH PhD student who recently completed her studies at SciLifeLab, and one of the study’s authors, makes a related point in the interview. Because the samples were collected longitudinally, the team could compare people to themselves. 

– It allowed us to see the protein levels from a person before an event such as vaccination or infection, Annika says, and we could also see how the protein levels might change afterwards.

In precision medicine, that may be one of the most important shifts of all. Instead of judging biology only against a broad population reference range, future healthcare may increasingly need to know what is normal for you – and how far a new measurement has moved from that personal baseline.

What the researchers found in the blood

The study offers several glimpses of what that could look like. The researchers used data-driven seroclustering to group samples into immune-response patterns that reflected infection and vaccination histories. They found that some anti-interferon autoantibodies – immune molecules that mistakenly target part of the body’s antiviral defense – were stably present over time and were associated with prolonged COVID-19 symptoms in some participants. They also mapped 664 protein quantitative trait loci, or genetic variants associated with protein levels, showing that some protein signals are anchored in genetics and remain comparatively stable over time. To better understand the biology behind the proteins, the researchers also drew on protein class annotations from the Human Protein Atlas, helping to distinguish among different blood cells and tissues and to explain why proteins could have ended up in these samples. And they identified transient changes associated with recent exposures, including the elevation of LAP3 after infection and TIMP3-related shifts after vaccination.  

From early signals to earlier intervention

Melén argues that medicine has underestimated the early origins of many chronic adult diseases. With today’s ability to profile proteins in tiny amounts of blood, he says, researchers have a chance to detect “the early signs of disease before it’s being established” – before the heart attack, before chronic lung disease, perhaps even before some cancers. 

The pediatric implications may be even more immediate. Melén notes that his group is already launching work in newborns using an extra drop of blood taken alongside the PKU sample. The idea is to measure proteins that may indicate a child’s later risk of lung disease and, eventually, help guide interventions.

– In a pediatric setting, where we have infants and children who definitely do not like blood sampling, a drop of blood from the finger is doable, Erik says. 

– We are still early in really implementing it as a routine approach, Schwenk says. 

The recent pandemic, in that sense, drove us to test new concepts: what can molecular health monitoring do when people cannot or should not travel to clinics? But he also points to the next question – whether repeated microsampling could track how someone responds to a medication, or how asthma and allergy phenotypes change across pollen seasons. 

A model for closer integration between research and care

There is another reason this study deserves attention in Stockholm beyond the science itself. It hints at a way of organising translational research and interactions that the region has long discussed: closer integration between hospital-based clinicians and technology-heavy research environments, such as SciLifeLab. Melén welcomes the idea of a more formal SciLifeLab presence inside hospitals, arguing that Stockholm could be at the forefront of integrating genomics and biomarker research into bedside use. Kere and Bendes make a more personal version of the same point. For scientists in training, they say, translational collaboration means being able to learn across disciplines, interpret results more confidently and move more naturally between the clinic, the wet lab and computational analysis. In the best sense, this is not just a collaboration model for a single paper. It is a training model for the next generation of precision medicine researchers.

Text: Gustav Ceder, Communications Officer at SciLifeLab, freelance journalist and communications consultant.