AI can Help Detect People at Risk of Developing Pancreatic Cancer

Dr Ananya Malhotra, research fellow in statistics, London School of Hygiene &Tropical Medicine, London, UK. Image Credit: ESMO Press Office.

Early detection of pancreatic cancer is very important to save lives. Artificial intelligence (AI) could enable earlier detection of such cancers.

The ability of AI has been described in a study that will be presented at the ESMO World Congress on Gastrointestinal Cancer, from July 1st to 4th, 2020.

Pancreatic cancer affects 12 in every 100,000 individuals, meaning that it would not be efficient to screen everyone and would subject many individuals to unwanted tests and possible side effects.

However, around 70% to 80% of patients are diagnosed only when cancer reaches the advanced phase when it is too late for any therapeutic treatment, and five years following diagnosis, only 6% of the diagnosed patients survive.

Hence, screening can enable early detection of cancer, when treatments are known to be the most effective, thereby enhancing the survival rates. Screening involves two major requirements. Firstly, a screening test that can be conducted easily has certain side effects. Secondly, a defined set of individuals who would gain most from the screening procedure because they are at higher risk of developing pancreatic cancer.

For instance, women aged 50 to 71 years undergo mammography as part of the breast cancer screening procedure. AI might be the urgently needed solution to define a set of people who are at higher risk of developing pancreatic cancer and would gain from the screening procedure; particularly as some recently potential results have denoted that non-invasive tests meant for pancreatic cancer may soon be available for patients.

People who develop pancreatic cancer are known to consult their general practitioners (GPs) regarding non-specific symptoms, like back pain or gastrointestinal problems, more often in the months and years before diagnosis when compared to their peers in whom pancreatic cancer does not manifest.

At an individual level, such symptoms may not probably trigger additional analyses for cancer. Now, scientists have developed a concept that AI may identify a combination of such non-specific symptoms associated with a higher risk of getting pancreatic cancer, which GPs would find it hard to detect.

Electronic health records from GP practices in the United Kingdom were used in this preliminary research. The analysis involved 1378 individuals aged 15 to 99 years and diagnosed with pancreatic cancer between 2005 and 2010.

Every patient was matched by gender and age to four individuals who did not contract the disease. Data on drugs, diseases, and symptoms in the two years before diagnosis were used to develop a model that predicts who is likely to contract pancreatic cancer.

We used AI to study a large volume of data and look for combinations that predict who will develop pancreatic cancer. It’s not possible for the human eye to recognise these trends in such large amounts of data.

Dr Ananya Malhotra, Study Author and Research Fellow in Statistics, London School of Hygiene &Tropical Medicine

The pilot research discovered that in individuals aged under 60 years, the new model could predict people who were at greater risk of developing pancreatic cancer up to 20 months prior to diagnosis.

Our model has estimated that around 1,500 tests need to be performed to save one life from pancreatic cancer. This is unlikely to be small enough to make screening viable just yet. However, it shows that AI holds potential to narrow down the number of people we need to screen.

Dr Ananya Malhotra, Study Author and Research Fellow in Statistics, London School of Hygiene &Tropical Medicine

Dr Malhotra continued, “We should be able to reduce this quite a lot further by matching pancreatic cancer patients to controls from the general population, which is what we plan to do next (in the current study, the controls had other types of cancer).”

Pairing this predictive model with a non-invasive screening test, followed by scans and biopsies, could lead to earlier diagnosis for a significant proportion of patients and a greater number of patients surviving this cancer,” Malhotra added.

Applying AI to detect individuals who are at higher risk of developing pancreatic cancer up to 20 months earlier could mean the difference between life and death.

This should be enough time to screen for pancreatic cancer, then proceed with diagnosis and treatment in patients with a positive screening test. Early diagnosis in pancreatic cancer gives the highest chance of cure.

Dr Angela Lamarca, Consultant in Medical Oncology, The Christie NHS Foundation Trust

GPs can apply this kind of AI model on their medical records to emphasize patients who are at greater risk of contracting cancer. As such, they could raise an alarm to show which individuals should receive screening, added Lamarca.

“We need bigger studies incorporating AI tools into daily clinical practice and exploring the benefit of screening the patients selected by AI. More research is also needed to find a good screening test for these high-risk patients,” Lamarca concluded.

This study was financially supported by the Pancreatic Cancer Research Fund.

Journal Reference

Malhotra, A., et al., (2020) Can we screen for pancreatic cancer? Identifying a sub-population of patients at high risk of subsequent diagnosis using machine learning techniques applied to primary care data. Annals of Oncology.


Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.