AI helps to identify new links between the pill and lower ovarian cancer risk

The contraceptive pill could help to reduce the risk of ovarian cancer in one in four women, AI-driven research suggests.

A new study which used AI to screen for risk factors of ovarian cancer, found that the oral contraceptive pill reduced the risk of the condition by 26 per cent among women who had ever used the pill, and by 43 per cent for women who had taken it after the age of 45.

Some biomarkers associated with ovarian cancer risk were also identified in the study, including several characteristics of red blood cells and certain liver enzymes in the blood.

The researchers from the University of South Australia (UniSA) also found that women who had given birth to two or more children had a 39 per cent reduced risk of developing ovarian cancer compared to those who had not had children.

UniSA researcher Dr Amanda Lumsden said: “Ovarian cancer is notoriously diagnosed at a late stage, with about 70 per cent of cases only identified when they are significantly advanced.

“Late detection contributes to a survival rate of less than 30 per cent over five years, in comparison to more than 90 per cent for ovarian cancers that are caught early. That’s why it’s so important to identify risk factors.

“In this research, we found that women who had used the oral contraceptive pill had a lower risk of ovarian cancer. And those who had last used the Pill in their mid-40s, had an even lower level of risk.

“This poses the question as to whether interventions that reduce the number of ovulations could be used as a potential target for prevention strategies for ovarian cancer.”

The study used AI to assess the data of 221,732 females (aged 37-73 at baseline) in the UK Biobank.

Machine learning specialist, UniSA’s Dr Iqbal Madakkatel, said the study showed how AI can help to identify risk factors that may otherwise have gone undetected.

“We included information from almost 3000 diverse characteristics related to health, medication use, diet and lifestyle, physical measures, metabolic, and hormonal factors, each measured at the start of the study,” Dr Madakkatel he said.

“It was particularly interesting that some blood measures – which were measured on average 12.6 years before diagnoses – were predictive of ovarian cancer risk, because it suggests we may be able to develop tests to identify women at risk at a very early stage.”