An overlooked organ may hold the key to understanding why some people suffer heart attacks despite having low cholesterol.
Cardiologists have spent decades managing blood pressure and lipid levels, yet coronary artery disease remains the leading cause of death worldwide. We know that circulating immune cells drive the chronic inflammation that ruptures arterial walls. Yet medicine has largely ignored the spleen, the primary training ground and reservoir for these cells, because it is too difficult to evaluate on standard clinical scans.
A new study changes this by using machine learning to analyze the spleen’s hidden architecture. The results suggest that the spleen is not just an innocent bystander in cardiovascular health. Instead, its physical texture may be a visible footprint of genetic cardiovascular risk, offering a new way to identify vulnerable patients.
The forgotten organ
Researchers used deep learning to analyze abdominal MRI scans from two large patient groups. They extracted 107 splenic radiomic features from 42,059 UK Biobank participants and 2,745 Mass General Brigham Biobank participants. This massive dataset allowed the AI to detect subtle patterns in tissue density and shape that escape the human eye, grouping patients by their splenic profiles.
The analysis revealed a direct link between the spleen and the heart. Here is what the data showed:
- Exactly 10 splenic features from the UK Biobank cohort were significantly associated with coronary artery disease.
- A genomic analysis of these features identified 219 genetic loci linked to spleen structure.
- The strongest link was found at 9p21, which is the most powerful genetic risk factor for coronary artery disease.
That genetic connection is the real story here.
A genetic breakthrough
For years, scientists have known that the 9p21 chromosome region increases heart attack risk, but they could never explain the underlying mechanism. This study suggests the answer lies in the spleen. Specifically, genetic variants at 9p21 were associated with splenic run-length nonuniformity, a measure of how heterogeneous the organ’s tissue texture is.
This finding complicates our understanding of heart disease. It suggests that genetic risk may manifest as physical changes in the immune system’s main reservoir long before plaque builds up in the arteries. If the spleen is actively driving vascular inflammation, treating the heart alone is merely addressing the symptom, not the source. It forces us to rethink cardiovascular risk as a systemic immune issue rather than a simple plumbing problem.
The translation gap
However, the transition from research lab to clinic will not be easy. While the AI model showed consistent results across the research databases, it faced significant hurdles when applied to routine clinical scans.
Real-world hospital MRIs suffer from highly variable imaging protocols and diverse patient populations. This clinical heterogeneity makes it difficult for the algorithm to reliably extract the same delicate splenic features. Until imaging standards are unified across hospital networks, this diagnostic tool remains confined to research pipelines rather than active triage.
Ultimately, this work proves that the path to understanding the heart runs through the abdomen. It challenges the cardiology community to look beyond the coronary arteries and begin studying the systemic immune organs that fuel arterial disease.
Read the full study in Science Translational Medicine.
