AI and proteomics uncover seven blood proteins linked to vasculitis remission

Vasculitis is a rare immune system disease that causes blood vessels to become inflamed, potentially damaging organs like the kidneys and lungs if left untreated. Physicians often rely on immunosuppressive medications to manage the condition, but these drugs can cause serious side effects. Identifying patients in remission is crucial to minimize unnecessary treatment and optimize regimens.

A Berlin research team has discovered that seven specific proteins in patients' blood plasma can reliably signal remission. Using proteomics—a technique that studies proteins—and artificial intelligence, they identified these biomarkers. The method could potentially be applied to other immune and inflammatory diseases, improving treatment precision while reducing side effects.

Dr. Philipp Mertins, head of the Proteomics Technology Platform at the Max Delbrück Center and BIH, led the collaboration between multiple institutions. He explained that existing biomarkers like CRP cannot accurately determine remission. The team identified thousands of proteins whose levels varied between patients in remission and those still active. After removing misleading proteins, they found seven that could reliably indicate remission through machine learning.

The discovery was validated in an independent group of patients, showing its potential in clinical practice. However, the causal relationships between these proteins and vasculitis remain unclear, and a widely available test for the assay is not yet available. The next step is to conduct larger clinical studies to determine the assay's effectiveness and benefits for patients.