Doctors grade heart-transplant rejection by looking at biopsy tissue under a microscope — but new research shows that similar-looking samples can hide very different biology, which may explain why some patients do not respond to standard treatment.
Researchers at Vanderbilt Health and the Translational Genomics Research Institute (TGen) applied spatial transcriptomics to repeated biopsy samples from 62 adult and pediatric heart-transplant recipients during and after rejection. Published August 10, 2026 in Nature Cardiovascular Research.
What spatial transcriptomics adds
The technique maps gene activity while preserving where each cell sits in the tissue — and that spatial element is the point.
Conventional gene-expression analysis requires grinding tissue up, which yields an average across every cell present and discards the architecture. For most questions that is acceptable. For rejection it is not, because rejection is fundamentally about spatial relationships: where immune cells have infiltrated, which structures they surround, whether they are clustered around blood vessels or dispersed through muscle.
A biopsy showing immune cells concentrated around vessels represents something different from one showing the same number spread through the tissue. Averaging erases exactly that distinction.
How rejection is graded now
Current practice is a pathologist examining stained tissue and assigning a grade based on how much immune infiltration is present and whether heart muscle cells are being damaged.
The system works and has genuine limitations. It relies on visual assessment, so grading varies between pathologists. It samples a tiny fragment of a large organ, so it can miss patchy rejection. And it categorises a continuous biological process into a handful of discrete grades.
Treatment follows the grade fairly mechanically — higher grades trigger intensified immunosuppression — which means anything the grade fails to capture is invisible to the treatment decision.
What the team found
They identified 28 immune and heart-cell types whose mix varied across rejection classes, and found substantial molecular differences within the same rejection grade — variation invisible to standard microscopy.
That within-grade variation is the central result. It means the grade is not capturing something clinically important, and it supplies a mechanism for an observation clinicians have long made: that patients with apparently identical rejection respond very differently to the same treatment.
The treatment-resistance signature
Patients who did not respond to standard anti-rejection therapy showed baseline T-cell hyperactivation and tissue-remodeling patterns.
Both components are informative. T-cell hyperactivation suggests the immune response in these patients is already operating at an intensity that standard immunosuppression cannot adequately suppress. Tissue remodeling indicates structural change is already occurring — the graft is not merely under attack but is being altered.
The word baseline carries the clinical weight. These features were present before treatment failure, meaning they were potentially detectable in advance rather than only recognisable in retrospect.
The chronic rejection link
Certain gene-expression signatures were linked to chronic rejection, specifically cardiac allograft vasculopathy.
That condition is the principal long-term threat to transplanted hearts: a progressive thickening of the coronary arteries within the graft that restricts blood supply. It is diffuse rather than focal, which makes it poorly suited to stenting or bypass, and it is a leading reason transplanted hearts eventually fail.
Because transplanted hearts are denervated, patients typically do not experience the chest pain that would normally signal restricted coronary flow — so the disease progresses silently and is often advanced when detected. Early molecular warning would therefore be genuinely valuable.
What it would change
Adding a molecular layer to the biopsy could sharpen diagnosis and prediction — distinguishing patients needing only standard care from those requiring alternative therapy, and flagging those at risk of long-term complications.
The practical appeal is that it uses tissue already being collected. Transplant recipients undergo surveillance biopsies on a defined schedule regardless, so this adds analysis rather than an additional procedure.
Why the pediatric inclusion matters
Including both adult and pediatric recipients in a single cohort is worth noting, because children transplanted for heart failure face a different long-term problem from adults.
An adult receiving a transplant in their sixties may reasonably expect the graft to last the remainder of their life. A child transplanted in infancy will almost certainly need a second transplant, and possibly a third, because grafts do not last indefinitely.
That changes the stakes of every rejection episode. Damage accumulated during a poorly controlled episode shortens the life of a graft that was already going to be outlived, and each subsequent transplant is harder — sensitisation from prior grafts narrows the pool of compatible donors, and repeat surgery is technically more difficult.
Preserving graft function in children therefore has compounding value, which is an argument for applying more sophisticated diagnostics in exactly the population where biopsy tissue is most limited and procedures are least tolerable.
It is a research finding, not a clinical test. Turning it into one requires validation in independent cohorts, a workable turnaround time and a demonstration that acting on the information improves outcomes. Research news, not medical advice.