AI tool accurately predicts tumor regrowth in cancer patients

• 

Doctors and scientists have created an artificial intelligence program that can reliably forecast the likelihood of tumor regrowth in cancer patients after therapy.

Clinical oncologists characterize the finding as "exciting," given that it has the potential to transform patient monitoring. While recent therapy improvements have increased survival prospects, there is still a potential that cancer will recur.

Monitoring patients following treatment is critical to ensuring that any cancer recurrence is addressed as soon as possible. Currently, clinicians must rely on traditional methods, such as those that focus on the initial amount and spread of cancer, to forecast how a patient will perform in the future.

A world-first study led by the Royal Marsden NHS Foundation Trust, the Institute of Cancer Research, London, and Imperial College London has developed a machine-learning – a type of AI – model that can predict the probability of cancer recurrence and outperforms existing methods.

“This is an important step forward in being able to use AI to understand which patients are at highest risk of cancer recurrence, and to detect this relapse sooner so that re-treatment can be more effective,” said Dr. Richard Lee, a consultant physician in respiratory medicine and early diagnosis at the Royal Marsden NHS Foundation Trust.

Less anxiety, better life

The OCTAPUS-AI study's principal investigator, Dr. Richard Lee, told the Guardian that it might be crucial in not only improving outcomes for cancer patients but also relieving their anxieties, with relapse "a primary cause of anxiety" for many. "We hope to push the boundaries to improve cancer patients' care, help them live longer lives, and lessen the impact of the disease on their lives."

The AI tool may result in recurrence being recognized earlier in high-risk individuals, ensuring they receive treatment more quickly, but it may also result in fewer unneeded follow-up scans and hospital visits for those at low risk.

“Reducing the number of scans needed in this setting can be helpful, and also reduce radiation exposure, hospital visits, and make more efficient use of valuable NHS resources,” Lee said.

In the retrospective study, clinicians, scientists, and researchers created a machine learning model to see if it could reliably identify patients with non-small cell lung cancer (NSCLC) who were at risk of recurrence after radiation. Machine learning is a type of AI that allows the software to predict outcomes automatically.

Lung cancer is the leading cause

Lung cancer is the major cause of cancer death worldwide, accounting for slightly more than a fifth (21%) of cancer fatalities in the United Kingdom. NSCLC accounts for approximately five-sixths (85 percent) of all lung cancer occurrences, and the disease is often treatable if detected early. In the UK, however, more than a third (36%) of NSCLC patients have a recurrence.

The researchers fed their algorithm with clinical data from 657 NSCLC patients treated at five UK hospitals, as well as data on numerous prognostic markers to better predict a patient's probability of recurrence.

These comprised the patient's age, gender, BMI, smoking status, radiation intensity, and tumor features. The AI model was then utilized by the researchers to define patients as having a low or high risk of recurrence, how long they might experience until recurrence, and overall survival two years after therapy.

The tool was proven to be more accurate than traditional approaches in forecasting outcomes. The findings of the study, which was funded by the Royal Marsden Cancer Charity and the National Institute for Health Research, were published in the journal The Lancet’s eBioMedicine.

“Right now, there is no set framework for the surveillance of non-small cell lung cancer patients following radiotherapy treatment in the UK,” said study lead Dr. Sumeet Hindocha, a clinical oncology specialist registrar at the Royal Marsden and Imperial College London. “This means there is variation in the type and frequency of follow-up that patients receive … Using AI with healthcare data may be the answer.

“As this type of data can be accessed easily, this methodology could be replicated across different health systems.”

The study represents an "exciting first step" toward developing a tool to guide cancer patients' post-treatment surveillance on a national and worldwide scale, according to Hindocha.