The risk of lung cancer is variable among smokers. The level of lung cancer risk that justifies screening chest CTs in smokers has not been determined. One proposed strategy is to use a recently developed lung cancer risk prediction tool (Bach et al. JNCI 2003; 95:470) to identify screening cases as those with a 1% or higher predicted risk of lung cancer within 10 years. We applied this tool to a cohort of sporadically occuring lung cancer patients to determine what proportion of patients would have been screened given a 1% risk cutoff.
Smoking history was obtained prospectively in 221 pts with thoracotomy for curative resection of lung cancer; 199 were smokers. Predicted risk was calculated retrospectively from data from day of surgery for the 174 who were age 50 or greater, had at least 10 pack-years of smoking exposure, and for whom complete smoking history was obtained.
174 (79%) of 221 patients with resectable lung cancer (M/F=104/117) were studied. Distribution of predicted risk is given graphically. Median risk was 7% (M 8%, W 6%). 8% of patients diagnosed with lung cancer had a predicted risk of less than 1% within 10 years.CONCLUSIONS: If screening programs were to only target individuals with a predicted risk of lung cancer of 1% or higher in 10 years, very few cancers among smokers would be missed.
Applying a strategy of risk prediction to screening techniques in smokers for lung cancer may worth considering.
R. Barrera, None.