“AI tool to improve management of oropharyngeal cancer: detecting extranodal extension (ENE)”

In France, in 2018, there were nearly 5,000 new cases of oropharyngeal cancer, including 3,793 in men and 1,200 in women, according to Public Health France. This disease affects the tonsils, the soft palate, the back of the tongue and the throat. According to Canadian Cancer Society“cancerous tumor is a group of cancer cells that can invade and destroy nearby tissue. It can also spread to other parts of the body.”

Determine treatment for oropharyngeal cancer

Most oropharyngeal cancers are caused by infection with the human papillomavirus (HPV), also called human papillomavirusaccording to Canadian Cancer Society. HPV infections are very common because the virus is easily transmitted through skin contact with any part of the body that is infected.

The main types of treatment for curing oropharyngeal cancer are surgery, radiation therapy, and chemotherapy. But, to determine which one is the most suitable, it is essential to evaluate theextension extranodale (ENE). This term means that cancer cells located in a lymph node have spread to the surrounding tissues. THE lymph nodes are the organs that defend the body against cancer and are located throughout the body but especially in the neck.

ENE is often difficult to detect during examinations. Moreover, if the presence of ENE is detected after surgery, it remains a risk factor for cancer recurrence. Thus, researchers have developed an artificial intelligence (AI) tool to detect ENE. This proved to be more effective than four head and neck radiologists. The study was published in the journal The Lancet Digital Health.

The artificial intelligence algorithm more efficient than doctors

“If ENE is found after surgery, patients must still have a long course of chemotherapy and radiotherapy (…) which is associated with complications and reduced quality of lifeunderlines Benjamin Kann, one of the authors of the study. What’s important about this study is that it tested the algorithm in a very large clinical trial where patients (…) were supposed to have no ENE but yet ended up to have some.“This means that the doctors had predicted that the patients would not have ENE and they eventually did.

The researchers therefore tested their AI algorithm on this population. The goal was to see if the AI ​​was successful in predicting ENE or if, like radiologists, this technology would have discovered it only later. Result: The AI ​​discovered many ENEs with a high degree of accuracy, greater than the four head and neck radiologists.

Better anticipating ENE would make it possible to administer fewer treatments to patients and therefore improve their quality of life. Ultimately, if future clinical trials corroborate these initial results, doctors may soon be able to use this tool to identify patients’ risk of ENE and improve their management. According Public Health Francefor cases diagnosed between 2010 and 2015, the 5-year survival was 41% in men and 53% in women.