EndoPredict
Gene-expression signature in early ER-positive, HER2-negative breast cancer. It answers whether endocrine therapy alone suffices or whether chemotherapy belongs with it.
You receive two numbers. The 12-gene Molecular Score measures gene expression in the tumour: eight tumour-associated genes, three reference genes, one control gene for DNA contamination. The EPclin score adds tumour size and nodal status. It ranges from 0 to 7; at 3.3 runs the boundary between low and high risk.
The EPclin is what matters. In the ABCSG-6/8 cohorts — 1,702 patients, endocrine therapy only, followed for a median of 9.6 years — 62.6 percent fell into the low-risk group. Their risk of distant metastasis lay well below that of the high-risk group: hazard ratio 4.77 (95 % CI 3.37–6.67). In the premenopausal patients of a separate validation (n = 385) it was 3.58.
It becomes clinically interesting where score and marker diverge. In an analysis of 1,652 prospectively examined cases, EPclin and Ki-67 did correlate significantly. Even so, 118 of 449 patients with a Ki-67 above 20 percent were classified by EPclin as low risk — more than a quarter of those in whom the conventional marker would have advised chemotherapy.
This is not a contradiction but the whole point. The score measures something that grading and proliferation index do not capture. It does not replace them; it comes in addition.
Without tumour size and nodal status we cannot compute an EPclin, and the molecular score alone does not carry the decision. Please send both with the case — otherwise we have to ask, and the report is delayed by a day.
We issue the report in English, signed by a specialist in pathology. It contains both scores, the threshold and the patient's placement.
One limitation we prefer to name ourselves: for the 21-gene and the 70-gene signature there are prospective studies (TAILORx, MINDACT). For EndoPredict the evidence is predominantly prospective-retrospective. The cohorts are large and followed for a long time, but it is a different level of evidence — something you should learn from us rather than have to find out yourself.
- Filipits M, et al. Prediction of Distant Recurrence Using EndoPredict Among Women with ER+, HER2- Node-Positive and Node-Negative Breast Cancer Treated with Endocrine Therapy Only. Clin Cancer Res. 2019;25(13):3865-72. PMID 31064782
- Constantinidou A, et al. Clinical Validation of EndoPredict in Pre-Menopausal Women with ER-Positive, HER2-Negative Primary Breast Cancer. Clin Cancer Res. 2022;28(20):4435-43. PMID 36043530
- Jank P, et al. Comparison of risk assessment in 1652 early ER positive, HER2 negative breast cancer in a real-world data set: classical pathological parameters vs. 12-gene molecular assay (EndoPredict). Breast Cancer Res Treat. 2022;191(2):327-33. PMID 34783927
A conversation before the first case
Material, fixation, and which clinical details we need. Half an hour now saves queries later.