• Research

A drop of blood to detect strokes and head trauma

In neurological emergencies, every minute counts. Stroke treatment must ideally be administered within four hours of symptom onset; beyond this window, the chances of recovery decrease sharply. This urgency prompted Jean-Charles Sanchez, professor in the Department of Medicine and founder of the Faculty Centre for Translational Biomarker Research, to develop a test capable of distinguishing strokes from a single drop of blood. This breakthrough is the result of three decades of applied research on cerebral biomarkers.

Issue 55 - December 2025

Sanchez_article_NL_600x600.jpg

Jean-Charles Sanchez, a biochemist, began his career at the clinical chemistry laboratory of the Geneva University Hospitals (HUG). “Diagnostics fascinated me from the very start”, he recalls. “It was also the start of proteomics, the large-scale study of proteins, which allowed researchers to detect disease- or trauma-induced changes in blood proteins.” Combined with bioinformatics, this approach enables the simultaneous analysis of multiple proteins, opening the door to powerful diagnostic applications.

The brain, however, presents particular challenges. The blood–brain barrier limits access to cerebral biomarkers, and medical imaging already provides substantial information once the patient reaches the hospital. “In clinical settings, imaging remains the most precise diagnostic tool”, he explains. “But outside the hospital, being able to rapidly identify biomarkers of stroke or head trauma can save lives.” In pre-hospital environments, where imaging is unavailable, a simpler, portable diagnostic tool could significantly accelerate triage and care. The concept emerged: to create a test that could be used before hospital arrival, saving precious time and directing patients to the most appropriate facility.

A revolution in stroke triage

In 2014, Jean-Charles Sanchez founded the spin-off company ABCDx, supported by UNIGE and in collaboration with Joan Montaner from Vall d’Hebron Hospital in Barcelona, where teams were also investigating cerebral biomarkers. “Too many discoveries remain in laboratories due to a lack of industrial development”, he notes. “I had to learn about patent filings, industrialisation, and regulatory requirements, without which no test can reach the market. I underestimated the time and energy required.”

ABCDx has since developed a rapid test capable of detecting large-vessel strokes, the most severe form, caused by the obstruction of a major artery, directly in the ambulance. How does the test work? A drop of blood from the fingertip, the simultaneous measurement of three proteins, and a photograph taken with a smartphone. An artificial-intelligence algorithm analyses the image, quantifies each protein, and integrates clinical parameters such as blood pressure, age, and blood glucose. The result appears instantly.

“The impact is particularly strong in regions with long transport distances”, he explains. “If paramedics know the patient has a large-vessel stroke requiring thrombectomy, they can call a helicopter immediately instead of losing hours and alert the hospital to prepare the team.”

The strength of the device lies in combining multiple biomarkers and clinical data. “One protein alone cannot answer all diagnostic questions. It is essential to distinguish a large-vessel stroke from a small-vessel stroke, a cerebral hemorrhage, or other stroke mimics.” The test aims for 100% specificity: a positive result confidently signals a large-vessel stroke. False negatives remain possible, but the goal is not to replace imaging, it is to enable rapid, accurate pre-hospital triage.

Multiple and promising applications

For the past five years, ABCDx has been conducting multicenter clinical studies in Germany, Spain, and Switzerland. Beyond stroke, the biomarkers have proven valuable in head trauma. A US patent has just been granted for the adult traumatic brain injury application, and a multicenter study published in the Journal of Neurotrauma shows that rapid blood testing after a blow to the head in children could support pediatric emergency teams, alongside clinical scores and physical examination.

Collaborations are also underway in pulmonary embolism (PE), aiming to combine D-dimer and NTproBNP within an algorithm capable of ruling out or confirming PE outside the hospital. This approach would also allow severity stratification, identifying high-risk patients who require urgent transfer. “Our objective is pragmatic: accelerating treatment by getting each patient to the right place at the right time.”

Jean-Charles Sanchez’s trajectory reflects the challenges of translational research. “My work has always been applied and open. We publish our patents as soon as they are filed to avoid delaying scientific publications.” This requires a careful balance between intellectual property protection and knowledge dissemination. Further developments aim to extend the technology to other life-threatening emergencies where immediate diagnosis can transform outcomes.

How does the test work?

  1. A drop of blood is taken from the fingertip.
  2. The blood is applied to a test cassette.
  3. Three colored bands appear, each corresponding to a different protein.
  4. The paramedic scans the cassette’s QR code and photographs it with a smartphone.
  5. The AI algorithm analyses the image and quantifies the proteins.
  6. Clinical data (e.g., blood pressure, glucose) are integrated.
  7. The result appears instantly on the smartphone, indicating the stroke type and the appropriate hospital.
  8. The information is simultaneously transmitted to the receiving hospital so the medical team can prepare.

Jean-Charles SANCHEZ
Associate Professor,
Department of Medicine
& Faculty Centre for Translational Research in Biomarkers (CITB)

In the same issue

S'abonner à la newsletter
Sign up to the newsletter

S'abonner