The target of the investigation was a distinction that matters enormously in the clinic but is surprisingly hard to make: the difference between luminal A and luminal B breast cancer. Both subtypes are hormone receptor positive, meaning their cells carry estrogen or progesterone receptors that fuel their growth, and both are generally more treatable than triple-negative or HER2-enriched disease. But they behave differently. Luminal A tumors tend to grow more slowly, carry a better prognosis, and usually respond well to endocrine therapy alone. Luminal B tumors are more aggressive, more likely to proliferate rapidly, and often require chemotherapy in addition to hormone-blocking drugs. The conventional way to separate them is immunohistochemistry, a staining technique that measures receptors and a proliferation marker called Ki-67. The problem is that a biopsy samples only a fragment of the tumor, while molecular characteristics can vary dramatically from one region of a mass to another. A needle that lands in a quiet corner of an aggressive tumor can lead to an underestimated subtype and undertreated disease.

The research team, led by Hui Zhou and Daoyu Yang with corresponding author Xianchun Zeng, set out to capture what the biopsy cannot: the whole tumor. Their approach rests on multiparametric MRI, a protocol that combines three complementary ways of imaging the breast. T2-weighted imaging reveals tissue texture and water content, highlighting edema and necrosis. Diffusion-weighted imaging tracks how freely water molecules move, which reflects cellular density and membrane integrity — densely packed, rapidly dividing cells restrict water diffusion in characteristic ways. Dynamic contrast-enhanced MRI, the workhorse of breast imaging, watches a gadolinium contrast agent wash into and out of the tumor over time, mapping the chaotic vasculature that malignant growths build to feed themselves. Each sequence contributes a different window onto the tumor’s physiology, and each window contains patterns invisible to the human eye.

To convert those patterns into numbers, the researchers used PyRadiomics, an open-source software platform that computes hundreds of quantitative features from a defined region of interest. These features include first-order statistics describing the distribution of pixel intensities, shape metrics capturing the tumor’s geometry and margins, and higher-order texture measures that quantify how intensities are spatially arranged — whether the tumor is homogeneous or patchy, smooth or irregular at scales ranging from a few pixels to the entire lesion. A single MRI study can yield well over a thousand such features, far too many to feed directly into a classifier without the model simply memorizing the training data. The team therefore built a four-stage filtering pipeline: variance thresholding to discard features that barely varied across patients, Pearson correlation analysis to remove redundant measures that duplicated one another, an ANOVA F-test to keep only features statistically associated with the subtype labels, and finally L1-regularized logistic regression, a technique that shrinks unimportant feature weights to zero, with five-fold cross-validation to guard against overfitting.

One further statistical hurdle deserved special attention. In the study cohort of 130 patients with pathologically confirmed invasive ductal carcinoma, luminal B tumors outnumbered luminal A tumors 86 to 44 — a ratio that reflects real-world prevalence but that would bias a naive classifier toward always predicting the more common class. The researchers applied the Synthetic Minority Over-sampling Technique, or SMOTE, which generates synthetic examples of the underrepresented luminal A class by interpolating between existing cases in the high-dimensional feature space. Crucially, they applied SMOTE only within the training folds of the cross-validation procedure, ensuring that no synthetic data leaked into the validation sets and inflated performance estimates. It is the kind of methodological discipline that separates credible machine learning studies from the many that quietly overstate their accuracy.

With the feature set refined and the class balance corrected, the team compared nine machine learning classifiers head to head, ranging from classical algorithms like logistic regression and support vector machines to gradient-boosted decision tree ensembles. The winner was XGBoost, an algorithm that builds a cascade of simple decision trees, each one trained to correct the residual errors of its predecessors. XGBoost has become a favorite in radiomics research because it handles heterogeneous feature types gracefully, resists overfitting through built-in regularization, and consistently tops machine learning competitions on tabular data of exactly this kind. The final model fused the radiomics features with clinical variables, and its performance was assessed with the area under the receiver operating characteristic curve, or AUC, a metric that captures how well the model separates the two classes across all possible decision thresholds.

The results were strongest at the institution where the model was developed. Using data from Center 1, comprising 64 patients, the clinical-radiomics fusion model achieved an AUC of 0.8944, with a standard deviation of 0.0711 across the cross-validation folds — a level of discrimination that approaches the reliability of the immunohistochemistry it is meant to complement. External validation, the true test of any predictive model, followed at two hospitals the model had never seen. At Center 2, with 36 patients, the AUC was 0.8483, still comfortably in the range considered clinically useful. At Center 3, with 30 patients, performance declined to 0.7269, a drop the authors attribute to the smaller sample size and the inevitable differences in scanner hardware, imaging protocols, and patient populations between institutions. Even so, the model retained meaningful discriminatory power at all three sites, and pairwise DeLong tests confirmed that the multiparametric fusion model significantly outperformed models built on any single MRI sequence alone. Decision curve analysis, which quantifies the net clinical benefit of acting on a model’s predictions at various threshold probabilities, showed the fusion model offered superior net benefit compared with simpler alternatives.

Perhaps the most forward-looking element of the study is its commitment to transparency. Black-box predictions are a hard sell in oncology, where clinicians need to understand why a model reached its conclusion before trusting it with treatment decisions. To open the box, the researchers applied SHAP analysis, a technique rooted in cooperative game theory that assigns each input feature a contribution value for every individual prediction. The analysis revealed that features derived from diffusion-weighted imaging and dynamic contrast-enhanced MRI were the most influential predictors of subtype, and — critically — the importance rankings remained consistent across all three centers. That consistency suggests the model is latching onto genuine biological signal, such as the cellularity captured by diffusion imaging and the vascular permeability captured by contrast dynamics, rather than artifacts of a particular scanner or protocol. It also gives radiologists a concrete hypothesis about which imaging phenomena encode proliferation status, the biological variable that most cleanly separates luminal A from luminal B disease.

The authors are careful to position the model as a complement to immunohistochemistry rather than a replacement. Biopsy remains the gold standard, and no regulatory agency is likely to clear an imaging classifier to stand alone in subtype determination. But the model addresses a genuine weakness of the current standard: sampling error. Because radiomics features are computed across the entire tumor volume, they integrate information from every region of the lesion, capturing intratumoral heterogeneity that a core needle can miss. In practice, a model like this could serve as a second opinion at the time of diagnostic MRI, flagging cases where the imaging signature of the tumor disagrees with the biopsy result and prompting clinicians to consider repeat sampling or more aggressive treatment planning. It could also help stratify patients in regions or healthcare systems where immunohistochemistry panels are delayed or incomplete.

The study is not without limitations, and the authors acknowledge them. The cohort of 130 patients is modest by the standards of deep learning, and the luminal A group in particular is small, which widens the confidence intervals around the performance estimates. The retrospective design means the model was tested on patients whose scans were acquired under each institution’s routine protocols rather than a standardized imaging regimen, which cuts both ways: it makes the external validation results more realistic, but it also leaves open the question of how much performance could improve with harmonized acquisition. The decline in AUC at Center 3 is a reminder that generalization remains the central challenge of clinical machine learning. Still, the trajectory of the field is clear. As imaging datasets grow and multicenter collaborations mature, quantitative MRI combined with interpretable machine learning is moving steadily from the research lab toward the radiology reading room — and for the hundreds of thousands of women diagnosed with hormone receptor positive breast cancer each year, the prospect of a fuller, faster, and less invasive picture of their tumor’s biology is a development worth watching closely.

Subject of Research: Multiparametric MRI radiomics and machine learning for preoperative differentiation of luminal A and luminal B breast cancer subtypes

Article Title: Multiparametric MRI-based radiomics combined with machine learning for preoperative differentiation of luminal A and luminal B breast cancer: a multicenter study

Article References: Multiparametric MRI-based radiomics combined with machine learning for preoperative differentiation of luminal A and luminal B breast cancer: a multicenter study. (n.d.). https://doi.org/10.1186/s12880-026-02805-2

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02805-2

Keywords: breast cancer, radiomics, multiparametric MRI, machine learning, XGBoost, luminal A, luminal B, DCE-MRI, diffusion-weighted imaging, SHAP, multicenter validation, tumor heterogeneity

The research team, led by Pernilla Wikström and Anders Bergh, focused on a concept they call tumor-instructed normal tissue, or TINT. The idea is that tumors are not passive invaders but active communicators, secreting signals that reprogram neighboring cells to create a supportive environment for growth and invasion. To test how far this instruction reaches and what it depends on, the researchers analyzed gene expression in tissue samples from 44 men with unifocal prostate cancer who had undergone surgical removal of the prostate. For each patient, they sampled the tumor itself, benign tissue taken close to the tumor, and benign tissue taken at a distance from it, all preserved as formalin-fixed paraffin-embedded samples and profiled using Clariom D microarrays.

The results were striking. The transcriptome — the complete set of RNA messages produced by the cells — in tumor-adjacent benign tissue differed significantly from that in cancer-free biopsy control tissue. Even more revealing, the expression patterns in this benign tissue tracked three distinct variables: the distance from the tumor, the tumor’s grade as judged by its aggressiveness under the microscope, and the tumor’s molecular subtype. In other words, a pathologist looking at seemingly normal tissue could, in principle, read out molecular information about a tumor they might not even be sampling directly.

To dissect what these molecular changes meant functionally, the team used gene set enrichment analysis, a computational method that tests whether predefined groups of genes — so-called hallmark gene sets representing known biological processes — are coordinately upregulated or downregulated in a given sample. They compared hallmark pathway activity in both tumor and TINT tissue across tumor grade, contrasting high-grade with low-grade cancers, and across molecular subtype, contrasting Luminal B with Luminal A tumors, a classification borrowed from systems that categorize prostate tumors by their cellular lineage and proliferative character.

A clear pattern emerged. Hallmarks related to cell proliferation were enriched in both the tumor and the adjacent benign tissue of high-grade and Luminal B tumors, and the researchers confirmed this at the protein level: immunohistochemical staining for Ki67, a well-established marker of cell division, was increased in the benign tissue near aggressive tumors. Signs of increased metabolism also appeared in both tumor and TINT tissue associated with high grade and the Luminal B subtype. This suggests that the metabolic reprogramming long recognized as a hallmark of cancer cells extends into the surrounding normal tissue, which may be co-opted to support the tumor’s energy demands.

The immune landscape told a different and more nuanced story. Rather than a simple gradient fading with distance, certain immune response programs were uniquely enriched in the benign tissue of high-grade tumors and, separately, in the benign tissue of Luminal A tumors. Critically, these immune signatures were more pronounced in tissue taken close to the tumor than in tissue taken farther away, indicating that the tumor’s immunological influence on its surroundings is both subtype-specific and spatially graded. This spatial dimension matters, because it implies that the tumor actively instructs its immediate microenvironment rather than merely reflecting a whole-organ field effect.

Perhaps the most clinically provocative part of the study involved biopsies. The researchers compared gene expression in tumor-negative biopsy samples from men who were later diagnosed with high-grade prostate cancer — samples that had been declared cancer-free at the time — with tumor-negative biopsies from men who remained cancer-free, and with biopsies that did contain cancer. The later-cancer group showed enrichment of metabolic gene programs compared with the controls, even though no clear proliferative or immunological differences were detectable. This hints that subtle molecular changes in apparently cancer-free tissue might foreshadow the presence or future emergence of aggressive disease elsewhere in the prostate.

That possibility points toward a genuinely new class of biomarkers. Today, prostate cancer diagnosis relies on needle biopsies that sample only a tiny fraction of the gland, and a negative biopsy does not rule out cancer hiding elsewhere. If the molecular state of benign tissue in a biopsy can indicate that a high-grade tumor is present in unsampled regions, pathologists could gain a powerful adjunct to conventional assessment. The study’s authors suggest that TINT alterations hold potential as diagnostic and prognostic biomarkers, and the grade- and subtype-specific signatures they identified could eventually help stratify patients without requiring the tumor itself to be captured by the biopsy needle.

The work also reframes how scientists think about tumor-host interactions. Much of cancer research focuses on the tumor cell itself or on the stromal and immune cells immediately within the tumor mass. The TINT concept pushes the boundary outward, treating the entire neighborhood of the tumor as a dynamically negotiated territory. The finding that proliferation, metabolism, and immune programs in benign tissue vary systematically with tumor proximity, grade, and subtype suggests a structured dialogue between cancer and host tissue — one that could be interrupted. If specific signaling pathways mediate the instruction of normal tissue, blocking them might deprive tumors of support from their surroundings, opening new avenues for therapeutic intervention.

Caveats remain, as they do in any study of this kind. The cohort comprised 44 men with unifocal disease treated at a single institution, and the biopsy analysis, while suggestive, identified metabolic enrichment rather than definitive predictive signatures. The authors themselves frame the findings as a basis for further exploration rather than a finished clinical tool. Still, the message is compelling: the benign tissue around a tumor is not a silent bystander but a molecular archive of the tumor’s identity and behavior. Learning to read that archive could change how prostate cancer is detected, graded, and ultimately treated, turning the tissue that pathologists currently dismiss as normal into one of the most informative samples in the gland.

Subject of Research: Gene expression changes in tumor-adjacent benign prostate tissue in relation to tumor proximity, grade, and molecular subtype

Article Title: Gene expression profiles in benign prostate tissue reflect tumor proximity, grade, and molecular subtype

Article References: Wikström, P., Freyhult, E., Semenas, J., Lundholm, M., Josefsson, A., Bergström, S. H., & Bergh, A. (2026). Gene expression profiles in benign prostate tissue reflect tumor proximity, grade, and molecular subtype. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-09035-8

Image Credits: AI Generated

DOI: 10.1186/s12967-026-09035-8

Keywords: prostate cancer, tumor-instructed normal tissue, TINT, gene expression profiling, transcriptomics, tumor microenvironment, molecular subtype, Luminal A, Luminal B, biomarkers, Ki67, gene set enrichment analysis