medwireNews: The Pan-Canadian Early Detection of Lung Cancer (PanCan) model, a low-cost risk prediction tool, effectively selects individuals at high risk for developing lung cancer, prospective study data show.
The model, which is a precursor to the validated PLCOm2012 model, incorporates data on age, education, family history of lung cancer, smoking duration, pack–years smoked, BMI, number of chest x-rays in the past 3 years, and history of chronic obstructive pulmonary disease to select candidates for lung cancer screening.
To investigate its efficacy, Stephen Lamb (British Columbia Cancer Agency, Vancouver, Canada) and colleagues screened 7044 individuals aged 50–75 years, who had smoked at some point in their life, for lung cancer risk using the PanCan model. Of these, 2537 with at a 2% or higher 6-year risk for lung cancer were enrolled into the study.
All study participants underwent low-dose computed tomography (CT) screening at baseline, and at 1 and 4 years later.
During a median follow-up period of 5.5 years, 172 lung cancers were diagnosed in 164 participants, giving a cumulative incidence of 6.5% and an incidence rate of 138.1 per 10,000 person–years.
There were 10 interval lung cancers detected; one was diagnosed between baseline and year 1 and the remainder occurred between years 1 and 4. On retrospective review, three (30%) were identified as nodules in the baseline scan.
The researchers therefore calculated that the low-dose CT screening had a sensitivity of 92.7% for detecting lung cancer.
When the study participants were stratified according to the presence of lung nodules at the baseline screen, Lamb and team found that those without nodules had a significantly lower lung cancer risk than those with nodules, with respective cumulative incidences of 0.18% (n=1) versus 4.6% (n=85) during 2 years of follow-up from baseline.
Lamb et al also point out in The Lancet Oncology that the cumulative incidence they observed in the PanCan study was significantly higher than the 4% reported the National Lung Screening Trial (NLST), which used age, pack–year, and smoking quit-time to predict risk.
In addition, the proportion of early-stage (I or II) lung cancers detected in PanCan was higher than in NLST, at 77% versus 57%.
Based on their findings, the authors conclude that the PanCan approach of using a highly predictive risk model rather than NLST-like criteria “should be considered for adoption in lung cancer screening programmes.”
In an accompanying comment, Harry de Koning, from the Erasmus Medical Center in Rotterdam, the Netherlands, says that the study “substantiates the ad-hoc analyses based on NLST that show superiority of using risk prediction models over the more classic risk factors— namely, age and smoking history alone.”
By Laura Cowen
medwireNews is an independent medical news service provided by Springer Healthcare. © 2017 Springer Healthcare part of the Springer Nature group