Clinical Software Effectively Categorizes Benign and Malignant Spots on Smokers’ Lung Scans

A study has developed a new clinical risk calculator software that effectively classifies, nine out of 10 times, which spots (lesions[nodules]) are benign and cancerous on an initial lung computed tomography (CT) scan among individuals at high risk for lung cancer.

The findings are expected to have immediate clinical bearing worldwide among health professionals who currently diagnose and treat individuals at risk for or who are diagnosed with lung cancer, and provide new evidence for developing and improving lung-cancer screening programs. A total of 12,029 lung cancer nodules seen on CT scans of 2,961 current and former smokers were studied in the population-based study.

The investigators, from the Terry Fox Research Institute (TFRI; Vancouver, BC, Canada), and other Canadian institutions, published their study’s findings in the September 5, 2013, issue of the New England Journal of Medicine (NEJM), and will have an instant impact on clinical practice, according to co-lead investigator Dr. Stephen Lam, chair of BC’s Provincial Lung Tumor Group at the BC Cancer Agency and a professor of medicine at the University of British Columbia (Vancouver, BC, Canada).

“We already know that CT screening saves lives. Now, we have evidence that our model and risk calculator can accurately predict which abnormalities that show up on a first CT require further follow-up, such as a repeat CT scan, a biopsy, or surgery, and which ones do not. This is extremely good news for everyone—from the people who are high risk for developing lung cancer to the radiologists, respirologists, and thoracic surgeons who detect and treat it. Currently, there are no Canadian guidelines for us to use in clinical practice.”

In countries where guidelines do exist, they largely relate to nodule size. The pan-Canadian investigators’ prediction model, developed by Brock University (St. Catharines, ON, Canada) epidemiologist Dr. Martin Tammemägi, includes a risk calculator that takes into account several factors in addition to size: older age, female sex, family history of lung cancer, emphysema, nodule location in the upper lobe, part-solid nodule type, lower nodule count and spiculation. “Reducing the number of needless tests and increasing rapid, intensive diagnostic workups in individuals with high-risk nodules are major goals of the model,” said Dr. Tammemägi.

The TFRI team used two sets of data to determine their findings, studying 12,029 nodules from 2,961 individuals—current and former smokers, aged 50–75, who had undergone low-dose CT screening. One set involved participants in the TFRI Pan-Canadian Early Detection of Lung Cancer Study from 2008 to 2010, where 1,871 individuals with a total of 7,008 nodules (102 of which were malignant) were screened and tracked. The other set involved 1,090 persons with 5,021 nodules (of which 42 were malignant) who took part in several lung cancer prevention trials conducted by the BC Cancer Agency during 2000-2010 and were funded through the US National Cancer Institute (NCI; Bethesda, MD, USA). In the former study, participants were followed for an average of three years; in the latter, for an average of eight-and-a-half years.

Dr. Lam noted that the prediction model holds up even in cases where clinicians are faced with the roughest challenges; for example, deciding what to do when nodules are one centimeter (the approximate width of an adult thumbnail) or smaller. Whereas nodule size is one predictor of lung cancer, the largest nodule appearing on the CT was not necessarily cancerous. The pan-Canadian study researchers discovered that nodules located in the upper lobes of the lung carry an increased possibility of cancer. In both data sets studied, researchers found that where cancer was present, fewer nodules were found. This model will simplify the work involved, particularly for radiologists, in evaluating and assessing nodules on scans, as well as respirologists and thoracic surgeons who must make decisions about tests and treatment for their patients.

“An accurate and practical model that can predict the probability that a lung nodule is malignant and that can be used to guide clinical decision making will reduce costs and the risk of morbidity and mortality in screening programs,” wrote Dr. Lam and study colleagues in the article.

“The findings in this study bolster the potential for the successful implementation of a lung cancer screening program using low-dose computed tomography [CT] within a high-risk population. This tool, combined with CT-screening, will increase our success in earlier detection, diagnosis and treatment of the disease. Further, this model combined with new guidelines for best clinical practice, will provide our health care system with both effective and affordable tools to implement such a program,” said thoracic surgeon Dr. Michael Johnston, a member of the study team from Nova Scotia (Canada). Dr. Johnston serves on the executive of the Terry Fox Research Institute and is chair of the medical advisory committee of Lung Cancer Canada.

“Many jurisdictions throughout the world are now considering whether or how to best implement lung cancer screening. Studies like this one are key to answering important questions so decisions are most likely to result in good practice and planning, and ultimately benefit patients,” said Dr. Heather Bryant, vice-president, cancer control at the Canadian Partnership Against Cancer.

The significant findings come on the heels of the US National Lung Screening Trial (2011) that found a 20% reduction in lung cancer mortality with the use of low-dose thoracic CT.

Dr. Christine Berg, co-principal investigator of the US National Lung Screening Trial and former chief, Early Detection Research Group, division of cancer prevention, for the National Cancer Institute in the United States, said, “This important work of Dr. Lam and colleagues is a major advance for clinicians performing lung cancer screening. They provide a tool to grapple with the problem of the high rate of positive low-dose CT scans. Fewer follow-up scans with their attendant cost and fewer biopsies with their complications will need to be performed while continuing to diagnosis lung cancer at an early stage to lower mortality. Coupled with continued public health efforts to lower cigarette smoking, this work will have international impact on the leading cause of cancer death worldwide.”

Related Links:
Terry Fox Research Institute
Brock University
University of British Columbia