In an earlier investigation, we did not detect a major genetic component to lung cancer in families of nonsmoking lung cancer probands. However, heterogeneity with respect to familial aggregation, based on probands' age at diagnosis, was evident. We reanalyzed our previously collected data of 257 families, stratified by age at diagnosis of the probands, using complex segregation analysis. We specifically tested the effects of a Mendelian diallelic gene, history of tobacco use, and history of selected chronic lung diseases in families with a proband diagnosed at the age of 60 years or older and in families with a younger proband (i.e. , under 60 years of age). Cases were identified from the Metropolitan Detroit Cancer Surveillance System. Information on lung cancer occurrence, smoking history, and chronic respiratory diseases in first-degree relatives was obtained for 210 older probands and for 47 younger probands. In older probands' families, no evidence of a major genetic effect was detected. A history of emphysema and tobacco-smoke exposure were found to be significant risk factors. In younger probands' families, a Mendelian codominant model with significant modifying effects of smoking and chronic bronchitis best explained the observed data. Our results suggest the presence of a high-risk gene contributing to early-onset lung cancer in a population where the probands are nonsmokers.
Copyright 1999 Wiley-Liss, Inc.