According to the authors of this
www.sciencedirect.com/science/arti...,
the difference between the prediction provided by PREDICT breast (breast.v3.predict.cam) and that provided by the ludicrously expensive OncotypeDx is that ....

4/4 CoI my group developed PREDICT breast. It is free to use and so those using an expensive test like OncotypeDx ought to have decent evidence of superior performance.

I know of only 1 publication that has evaluated the improvement in performance achieved by adding the OncotypeDx RS to PREDICT breast showed a trivially small improvement in discrimination and a tiny reclassification. Hard to justify its cost on that basis.

This total lack of understanding of the basics of risk prediction explains how the manufacturers (and investors) of OncotypeDx can pull the wool over the eyes of so many doctors and persuade them that their test is worth spending thousands on (PREDICT breast is free).

3/n.... should also be provided. Based on Fig 3D calibration of new model and OncotypeDx appears poor - but hard to be sure. Moreover, why not compare the performance of these models with a well-established model - PREDICT breast (breast.v3.predict.cam) that has been carefully validated?

6/
Gene expression was a breakthrough 20 years ago.

Despite academic ubiquity, only OncotypeDx, MammaPrint, and ProSigna actually improve breast cancer decisions.

Wrong data choice kills your model before training starts.

A 10-gene signature can now predict recurrence risk for HR+/HER2- early breast cancer #EBC patients post-chemo-endocrine therapy.

@myesmo.bsky.social @breastcancerorg.bsky.social

#Onco404 #Cancer #Kanser #BreastCancer #MemeKanseri #OncotypeDX #MammaPrint #EndoPredict #Prosigna #PAM50 #GGI #MedSky

🔍 Why It Matters
☑️ 20% of HR+/HER2- EBC patients relapse despite adjuvant C-ET.
✅ Existing tools (e.g., 21-gene, 70-gene) guide chemo decisions but fail to pinpoint high-risk patients after C-ET.

#Onco404 #Cancer #Kanser #BreastCancer #MemeKanseri #OncotypeDX #MammaPrint #Prosigna #OncSky #MedSky

Was recently shown OncotypeDx report @OncotypeIQ for a breast patient. At least four inconsistencies in this one report: A) different risk groupings; B) no incorporation of clinical factors in risk estimates; C) 3 different endpoints; D) 95% CI without central estimate.

Adjuvant anthracycline- plus taxane-based chemotherapy regimen improves survival in high-risk breast cancer patients

www.news-medical.net/news/2024121...

Prediction of OncotypeDX recurrence score using H&E stained WSI images https://www.medrxiv.org/content/10.1101/2025.07.21.25331907v1