Integrating digital pathology with genomic and clinical data may enhance prediction of late recurrence and improve selection of adjuvant therapy intensification in HR-positive HER2-negative disease.
Digital pathology is a “very busy field of research” today, the professor said. Peter Dubsky Several summaries were discussed at Rapid Oral Session 1 at the ESMO Breast Cancer 2026 Conference, University of Lucerne, Switzerland. “Over the past three years, we have produced high-impact publications based on a very general methodological framework that starts with scanning images. H&E slidecomplement the information it contains with other data. The result is prognostic score This can be used, for example, to simply extrapolate very expensive data. gene expression test Treat breast cancer (Lancet Oncol. 2026 Apr;27(4):512-526) or predict distant recurrence and optimize patient selection for CDK 4/6 inhibition (J Clin Oncol. 2025 Oct;43(28):3090-3101). ”
In addition to this growing body of research, data presented in Congress digital pathology (AI-Path) Predict 10-year distant recurrence risk using the OncotypeDX21 genetic recurrence score. HR-positive, HER2-negative early breast cancerRSClin N0 also incorporates clinical tumor characteristics and is superior to existing risk stratification tools and can improve selection of patients who may benefit from intensification of adjuvant therapy (Abstract 3RO).
The study included 6,319 patients with HR-positive, HER2-negative, axillary lymph node-negative breast cancer. TAILORx trial (N Engl J Med. 2018 Jul 12;379(2):111-121) (4,233 for fitting, 2,086 for validation) and 979 patients from the Chicago cohort. AI-Path outperformed RS in TAILORx (C-index 0.678 vs. 0.640, p < 0.001) with better prediction of late DR after 5 years (C-index 0.650 vs. 0.568). The new biomarker, AI-PathClinRS, consistently outperformed RSClin (p=0.0006)C-index 0.710/0.730/0.794 in the TAILORx fitting (p = 0.006) and validation cohorts (p = 0.002) and identified approximately twice as many high-risk patients as AI-PathClinRS. NATALEE Eligibility Criteria (shape).
shape. Multimodal clinical, pathological, and genomic models (PathClinRS) outperform current risk stratification tools (ESMO Breast Cancer 2026, Abstract 3RO)
During the session, Dubsky emphasized the prognostic value of multimodal biomarkers. “What I find impressive is the late time point of recurrence, the prognostic performance, and the increase in C-index to almost 0.7, which is rarely observed with traditional gene expression biomarkers,” he said. It highlighted not only advances in research but also changes in research methods. digital biobank Currently recognized in oncology. “I believe fully digital biobanks will soon become commonplace, but that’s something we need to prepare for. As we continue to integrate additional data sources such as circulating DNA, imaging modalities, and spatial biology, we will be able to generate true data.” Individual prediction of recurrence” he concluded.
