Introduction: Familial clustering of lung cancer has long been noted; multiple genetic analyses of lung cancer pedigrees have been performed in search of lung cancer predisposition genes. Because smoking is also observed to cluster in families, this major environmental risk factor can confound any evidence for a genetic contribution to lung cancer. We show evidence that a genetic contribution appears likely to contribute to non-smoking-related lung cancer.

Data and Methods: Using a population-based resource consisting of the computerized genealogy of the Utah pioneer founders and their descendants linked to a statewide cancer registry from 1966, we have analyzed the clustering of lung cancer cases among close and distant relatives. Using linked death certificate data that includes data on the contribution of tobacco to death, the subset of deceased lung cancer cases was stratified into smoking- (n=1895) and non-smoking-related (n=784) subsets. A test for a significant excess of pairwise relatedness among all cases was performed by comparing the average relatedness of these two independent subsets of lung cancer cases to the expected relatedness in the Utah population based on analysis of matched control sets (the well-published GIF method was used).

Results: Significant evidence for overall excess pairwise relatedness was observed for the 784 lung cancer cases whose death certificate stated that tobacco did not contribute to death (p= 0.001). These results suggest familial clustering, but cannot discriminate between genetic- or environmental-based clustering. To remove any effect of a familial smoking effect that is not genetic, all pairwise relationships closer than third-degree were ignored; significant excess relatedness was still observed among non-smoking cases (p=0.019). When a similar analysis was performed for lung cancer cases whose death certificate indicated a contribution to death from tobacco (n=1895), overall significant excess relatedness was observed (p<0.001), suggesting familial clustering of cases; however, when close relationships were ignored there was no longer evidence for significant excess relatedness (p=0.271).

Conclusions: Using published methods to analyze a powerful resource linking cancer data to genealogy we have shown that while lung cancers do appear to cluster in families, there are two distinct subsets of cases. The familial clustering of those cases who do not smoke provides strong evidence for a genetic contribution, while the familial clustering of those lung cancer cases who smoke appears to resemble the clustering expected with the shared environmental risk factor of smoking. Gene identification efforts should focus on the subset of lung cancer pedigrees representing cases for whom smoking did not contribute to lung cancer.

Citation Format: Lisa A. Cannon-Albright, Shamus Carr, Wallace Akerley. Evidence for a genetic contribution to non-smoking-related lung cancer. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1298. doi:10.1158/1538-7445.AM2014-1298