Machine Learning Assisted Oncotype Dx Recurrence Risk Classification Forecast System for Breast Cancer Based on 111,635 Patients Covering 11 Years

Background: TAILORx data confirm that using Oncotype DX (ODX) to assess the risk of early-stage breast cancer recurrence can spare women unnecessary chemotherapy. However, high up-front costs (list price, $4175) could dissuade usage. Also, from a technical perspective, this test cannot be widely used in developing countries, especially in relatively poor areas. Methods: By analyzing the Surveillance, Epidemiology, and End-Results database, multinomial ordinal logistic regression, random forest, gradient boosting decision tree, extreme gradient boosting, and multi-layer perceptron were adopted to develop an assistant forecast system for recurrence risk—low to intermediate (RS=2–25) or high (RS=26–100)—based on individuals’ sociodemographic and clinicopathological information. We validated the system using an independent dataset (training-test split of the original dataset) as validation. Results: We identified 111,635 patients with breast cancer: 86,617 patients (77.59%) were aged 50 years or younger; 23,514 patients (21.1%) had low risk scores, 71,439 patients (64.0%) had intermediate risk scores, and 16,682 patients (14.9%) had high risk scores. Variables closely associated (all P<0.05) with RS included age, sex, race, primary tumor site, histopathological grade, tumor size, pathology, and progesterone receptor status. In validation, for the multi-layer perceptron model, overall accuracy was 86.1% (sensitivity=0.69, specificity=0.87, and area under the curve=0.80).Conclusions Our assistant forecast system, based on sociodemographic and clinicopathological data, provides an alternative tool to estimate patients’ breast cancer recurrence risk. Funding: This work was supported by the Natural Science Foundation of China (No. 81872160), the Natural Science Foundation of China (No. 82072940), the China National Key R&D (or Research and Development) Program (No. 2020AAA0105000 and 2020AAA0105004), the Beijing Municipal Natural Science Foundation (Key Project) (No. 7191009), the Beijing Municipal Natural Science Foundation (No. 7204293), the Special Research Fund for Central Universities, Peking Union Medical College (No. 3332019053), the Beijing Hope Run Special Fund of Cancer Foundation of China (No. LC2019B03), the Beijing Hope Run Special Fund of Cancer Foundation of China (No. LC2019L07), the Beijing Hope Run Special Fund of Cancer Foundation of China (No. LC2020L01), the Golden Bridge Project Seed Fund of Beijing Association for Science and Technology (No. ZZ20004), and the National Key R&D Program of China (No. 2018YFC1315000 and No. 2018YFC1315003). Declaration of Interest: None to declare. Ethical Approval: The analysis followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College institutional review board approved this study as exempt.