One of life's serendipitous moments brought Chris Douglas to Dr. Stephen Lam.
Douglas, who lives in Vancouver, had been a smoker for more than 40 years. Lam, a renowned lung cancer specialist with the B.C. Cancer Agency, was working on part of a widespread research trial aimed at giving doctors an early, accurate and reliable way to determine which spots on lung scans are likely cancerous, and which are harmless.
It was 2008. Douglas's wife saw a newspaper article about Lam's trial and how they were looking for smokers to volunteer. Soon after, Douglas was enrolled and receiving annual computed tomography (CT) scans to look for cancerous growths in his lungs.
Sure enough, in 2011, Lam found just such a growth in Douglas's lungs, and that early detection enabled Douglas to have early, life-saving surgery.
Because of such success stories, there is hope that Lam's research will lead to important changes in the way CT scans are used in lung cancer detection, leading to early diagnoses when life-saving treatment is still possible, while saving other patients from unnecessary and costly follow-up treatments.
The results of the research, published Thursday in the New England Journal of Medicine, suggest that the tool developed by the researchers can accurately detect when lung nodules are cancerous or benign more than 90 per cent of the time. Many researchers believe this work could represent an important development in the early diagnosis and treatment of lung cancer, which is the deadliest type of cancer in Canada. According to the Canadian Cancer Society, 25,500 Canadians will be diagnosed with lung cancer this year and 20,200 will die from the disease.
"It's a huge deal," said Heidi Roberts, a radiologist in the Princess Margaret Cancer Program at Toronto's University Health Network, who did not directly participate in this study. "It has a very high accuracy, it's easily available and it's easily applicable. I think it has huge potential to be used widely."
Doctors have long used CT scans to find nodules in the lungs. But there has not been a good, reliable way to spot the few cancerous nodules among the numerous benign ones that are commonly found in the lungs of patients. As a result, many people needlessly undergo further scans, biopsies and other treatments to rule out cancer, which can cause unnecessary harm and lead to huge costs. There is also an ongoing debate among medical experts and policy-makers over the use of CT scans – some believe scans are unreliable and cause too much harm to be justified for widespread use.
The software-based algorithm used in the research trial, developed by Brock University's Martin Tammemagi, may change that perception. The tool uses a number of variables, such as a patient's age, gender, nodule size and location, to assess whether it's cancerous. From other screening models that are used to help interpret the results of CT scans for lung cancer, Lam notes, about 20 per cent of patients will typically be referred for follow-up scans, biopsies or other treatments. With the new tool, the number of referred patients drops to about four per cent.
"That's a significant reduction in terms of cost and also potential complications from many biopsy procedures," said Lam, co-lead investigator of the study and chair of B.C.'s provincial lung tumour group at the B.C. Cancer Agency. He added that researchers used low-dose CT scans, which emit far less radiation than regular CT scans.
Douglas, who quit smoking, had one lobe from his lung removed before the disease had spread more widely and is now able to enjoy swimming and playing with his grandchildren.
"They saved my life," he said of Lam's team.
In Canada, there are no national screening guidelines for lung cancer, but Lam and his colleagues are working to change that. The Canadian Partnership Against Cancer has also set up a task force to determine how to implement cross-country screening.