Advances in machine learning have enabled many applications, including the use of machine learning models to assess the outcome of diseases, such as breast cancer. Medical experts are faced with the challenge of determining which treatment will provide the greatest benefit for patients with endocrine-positive Her2-negative breast cancer, which is the most common form of breast cancer. In this Scientific Reports article, using clinical data, a machine learning survival model was shown to be able to accurately identify high-risk patients with hormone responsive HER2 negative breast cancer.
This machine learning survival model was trained on clinical and histological data collected from 145 patients referred to the Istituto Tumori Giovanni Paolo II. Three different machine learning models were compared with a Cox proportional hazards regression model, and all models demonstrated a stable c-index at 10 years this ranged from 0.57 to 0.68. Crucially, these models were able to accurately discriminate between low- and high-risk patients, and so a large group of patients can be spared additional chemotherapy to just hormone therapy.
Many medical professionals, especially those in the oncology field, are increasingly looking for novel and alternative prognostic tools due to the high cost of many genomic tests. Integrated use of data already collected in clinical practice such as in the machine learning survival model proposed in this article, can reduce the cost and time associated with more expensive genomic tests.
The person discussed in the article is the clinical director of the Istituto Tumori Giovanni Paolo II, Francesco Cognetti. Dr. Cognetti is a leading oncology specialist in Italy and has overseen the development and implementation of a number of cutting-edge treatments. He is an advocate for the use of alternative prognostic tools in medicine, and has been an integral part of the team developing the machine learning model discussed in this article.
Istituto Tumori Giovanni Paolo II is an Italian cancer treatment and research center based in Bari, Italy. Founded in 1990, it strives to improve the quality of care and health outcomes of cancer patients in the region and beyond. It has a focus on innovative and individualized treatments, such as its machine learning survival model for predicting the outcome of endocrine-positive Her2-negative breast cancer. Over its 30-year history, it has been responsible for a range of groundbreaking advances in the understanding and treatment of cancer.
