Brock Epidemiologist develops lung cancer prediction model

lung cancer

A Brock professor has developed a lung cancer prediction model that has been shown to identify at-risk individuals. Doctor Martin Tammemägi has led a team of Canadian researchers in a study that recruited people to participate in a lung cancer screening using his prediction model. Combined with screening and follow-up, the study was successful in identifying cases of lung cancer, 75 per cent of which were caught in early stages.

Tammemägi’s model, PLCOm2012, works by aggregating data provided by an individual regarding age, race/ethnicity, education level, body mass index (BMI), history of chronic obstructive pulmonary disease (COPD), family and personal history of lung cancer, current smoking status, average lifetime smoking intensity, smoking duration and, in former smokers, how long ago they quit. The model produces a statistic for how likely it is that the individual will develop lung cancer within the next six years.

A study was published on October 18 in Lancet Oncology discussing the model, PLCOm2012, and the lung cancer screening. The Pan-Canadian Early Detection of Lung Cancer Study (PanCan Study) was undertaken to test the model’s effectiveness. 2,537 individuals, all previous or current smokers between the ages of 50 and 75 were selected for the study. Each of those individuals had a minimum of a two per cent chance of developing lung cancer within the next six years as identified by the PanCan model, which is a precursor to the PLCOm2012. The individuals selected were screened using low dose computed tomography (LDCT) at the beginning of the study, after one year and again after 4 years.

The screening revealed lung cancer in 164 individuals, comprising 6.5 per cent of those tested. Of those 164, 75 per cent of people diagnosed were in early, potentially curable stages (Stages one or two); this is a higher identification rate than alternative models have produced. A study conducted in the United States, the National Lung Screening Trial, demonstrated that LDCT could reduce lung cancer mortality rates by up to 20 per cent. Combined with the success of the PLCOm2012 at identifying high risk patients, this mortality rate could be further reduced.

“Lung cancer is the leading cause of cancer death in Canada,” says Tammemägi. “I wanted to help reduce the burden of lung cancer mortality. Combining my lung cancer risk prediction model with LDCT lung cancer screening is expected to advert many lung cancer deaths.”

The model is being used in a pilot project for LDCT screening. Individuals are chosen for screening based on a two per cent risk factor as identified by Tammemägi’s prediction model. Tammemägi explained that if the pilot project is successful, lung cancer screening may become available province-wide, and it will be covered by OHIP.

As for the model itself, Tammemägi hopes to do some more targeted research to further refine the model: “I plan to evaluate and refine the PLCOm2012 and related models using Canadian longitudinal data to determine whether subsets of the population are at particularly elevated risk and the prediction models need to reflect this.”

He hopes to make lung cancer screening more efficient through emphasizing important components of the model.

The province of Ontario is adopting PLCOm2012 into public healthcare practice, and there is potential for other provinces and jurisdictions to follow Ontario’s lead.

“My goal is to optimize adoption of the model, or an evolved version of it, into lung cancer screening programs, to reduce lung cancer deaths,” says Tammemägi. “In addition, one of my general research goals is to optimize lung cancer screening programs, and this includes optimal incorporation of smoking cessation programs into the screening program, because quitting smoking will additionally save many lives.”

Pin It

Leave a Reply

Your email address will not be published. Required fields are marked *

* Copy This Password *

* Type Or Paste Password Here *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>