A Brock University-designed lung cancer prediction model is more effective at identifying Indigenous individuals in the U.S. who should be screened for lung cancer than what is currently used in the country’s health-care system, according to a research paper published Oct. 9 in the journal Cancer.
Brock Professor Emeritus of Epidemiology Martin Tammemägi’s cancer prediction model identified up to 24 per cent more Indigenous and non-Indigenous people who needed lung cancer screening than the United States Preventive Services Task Force (USPSTF) criteria.
Low-dose computed tomography (CT) lung cancer screening for high-risk individuals detects earlier stages of cancer, which is often cured by surgery. Without screening, most lung cancers are detected at an advanced stage, with most cases having low survival rates.
“Compared to the USPSTF criteria, my model takes into consideration many more factors that quantify the probability someone will get lung cancer, which is the basis for identifying people at high risk who will benefit from screening,” says Tammemägi.
The USPSTF criteria uses age and smoking history to determine if someone is eligible for lung cancer screening: individuals ages 50 to 80 years who currently smoke or have quit within the last 15 years; and those who have smoked 20 ‘pack years,’ such as one pack a day for 20 years or two packs a day for 10 years.
“In general, the USPSTF criteria is simplistic and narrow,” says Tammemägi.
In contrast, there are 11 criteria in Tammemägi’s PLCOm2012 model that take into account an individual’s personalized situation, including: the number of cigarettes smoked per day, the number of years of smoking, race/ethnicity, age, highest level of education obtained, family history of lung cancer and body mass index, among others.
The model is accompanied by mathematical equations to predict the risk of an individual getting lung cancer.
Tammemägi and his team retrospectively reviewed medical records for 1,565 Indigenous and non-Indigenous individuals with lung cancer in South Dakota. They assessed their eligibility for lung cancer screening using the USPSTF 2013 and 2021 criteria and two versions of Tammemägi’s PLCOm2012 model, one that included the race category and the other — ‘PLCOm2012norace’ — without the race category.
Sixty-six per cent of the research participants were identified as being eligible for lung cancer screening using the USPSTF 2013 criteria. That number increased to 89.6 per cent using the PLCOm2012norace model and 90.7 per cent using the PLCOm2012 model that included the race category. Both Indigenous and non-Indigenous research participants had similar results.
The team tested two versions of the PLCOm2012 model to assess the impact of removing race as a risk predicter. Increasingly it’s being recognized that race is a social construct and may not be appropriate to assess risk for disease, says Tammemägi.
“Although race is a social construct, until the risk factors for this construct are identified and included in risk prediction models, jurisdictions with large populations of underserved ‘races’ who are found to be at excess risk — including many Indigenous populations — should consider using risk prediction models incorporating race as a predictor variable,” he says.
He says Indigenous populations tend to start smoking at an earlier age than broader populations, hence develop lung cancer at earlier ages. There are other risk factors that researchers are still trying to identify.
Tammemägi is currently part of a team studying how to better screen for lung cancer in research, funded by the Canadian Institutes of Health Research, looking at cancer screening in Ontario among First Nations, Inuit and Métis peoples, a study funded by Canadian Partnership Against Cancer (CPAC) in Inuit peoples in the Champagne region, as well as two studies of Māori populations in New Zealand and one study of Aboriginals and the Torres Strait Islanders in Australia.