Abstract #2741: A multi-gene test for lung cancer risk

Abstract

Background: Approximately 90% of lung cancer cases are associated with cigarette smoking yet only 10-15% of heavy smokers develop lung cancer. Inability to identify this subset hinders efforts to monitor for early detection. Methods: A test comprising transcript abundance levels of 14 antioxidant, DNA repair and transcription factor genes was measured in normal bronchial epithelial cell samples in two casecontrol studies including a training set of 49 individuals (25 cancer cases and 24 controls) and an independently collected validation set of 40 individuals (18 cases and 22 controls). Results: In the training set the receiver operating characteristic (ROC) area under the curve (AUC) for the test was 0.82 (95% CI, 0.68 - 0.91) and using the best cut-off the odds ratio for cancer in the group with a positive test was 12.8 (95% CI, 3.2 - 50.9; P<0.001). When applied to the validation set (N=40), the ROC AUC was 0.87 (95% CI, 0.73 - 0.96) and the odds ratio was 15.8 (95% CI, 3.3 - 74.3; P<0.001). For the combined training and validation sets (N=89), the ROC AUC was 0.84 (95% CI, 0.75 - 0.91) and the odds ratio was 13.9 (95% CI, 5.0 - 38.8; P<0.001). In the subset, age of 50 or over and 20 pack-years smoking history or greater, (N=49) the odds ratio was 8.17 (95%CI, 2.13 - 31.4; P=0.002). Conclusion: This test will enable identification of individuals with heavy smoking history who are genetically predisposed to lung cancer so that they may be more closely monitored.

Citation Information: In: Proc Am Assoc Cancer Res; 2009 Apr 18-22; Denver, CO. Philadelphia (PA): AACR; 2009. Abstract nr 2741.