PanCan Model Can Select High-Risk Individuals for Lung Screening

October 30, 2017

A new risk model was effective at identifying individuals subsequently diagnosed with early and potentially curable lung cancer.

A new risk model was effective at identifying individuals subsequently diagnosed with early and potentially curable lung cancer, according to a new study.

“Low-dose CT lung cancer screening is most effective when applied to high-risk individuals,” wrote study authors led by Martin C. Tammemagi, PhD, of Brock University in St. Catharines, Ontario. The criteria currently used to select for lung screening come from the National Lung Screening Trial (NLST), and include age between 55 and 80 years, a smoking history of at least 30 pack-years, and regular smoking within the last 15 years.

The Pan-Canadian Early Detection of Lung Cancer (PanCan) study assessed the effectiveness of a different risk production model, incorporating the following variables: age, smoking duration, pack-years, family history of lung cancer, education level, body mass index, chest x-ray in the past 3 years, and history of chronic obstructive pulmonary disease. The study enrolled 2,537 eligible ever-smokers to test this model and undergo low-dose CT screening, and followed them for a median of 5.5 years; the results were published in Lancet Oncology.

During the follow-up period, there were 164 individuals who received a diagnosis of lung cancer (172 total lung cancers), yielding a cumulative incidence of 6.5%, which was greater than the NLST’s rate of 4.0%; the incidence rate was 138.1 per 10,000 person-years.

The median PanCan model risk score in the full cohort was 0.033; the median score among 752 patients (30%) lost to follow-up was 0.0335, compared with 0.0326 in those who were not lost.

There were 137 surgical resections done in patients with lung cancer, and 88% of those were early-stage lung cancer and 12% were benign lesions. Of the 172 total lung cancer cases, 80% were identified at the initial screen, 5% at years 1 and 2, and 10% at the fourth year. There were also 10 interval lung cancer cases, defined as those diagnosed within 1 year of a negative screen. One of those was diagnosed between the baseline screen and year 1, and the other 9 between year 1 and year 4.

More patients were diagnosed with early-stage disease (stage I or II non–small-cell lung cancer or limited-stage small-cell lung cancer) in the PanCan study (77%) than in the NLST (57%; P < .0001).

The authors noted that the study was limited by its nonrandomized nature, which makes comparison with NLST criteria difficult. Still, they wrote, “The results of the study support the notion that selecting individuals for lung cancer on the basis of a highly predictive risk model is more effective than using NLST-like criteria. These results could have public health and clinical implications.”