No abstract

We carried out a genome-wide association study of lung cancer (3,259 cases and 4,159 controls), followed by replication in 2,899 cases and 5,573 controls. Two uncorrelated disease markers at 5p15.33, rs402710 and rs2736100 were detected by the genome-wide data (P = 2 × 10 -7 and P = 4 × 10 -6 ) and replicated by the independent study series (P = 7 × 10 -5 and P = 0.016). The susceptibility region contains two genes, TERT and CLPTM1L, suggesting that one or both may have a role in lung cancer etiology.We and others have recently reported a susceptibility locus for lung cancer in gene region 15q25, an area that includes a cluster of nicotinic acetylcholine receptor genes [1][2][3] . In order to identify further susceptibility gene loci, we genotyped an additional 1,291 cases and 1,561 controls from three further studies (Toronto case-control study,

Using a model-based approach, we estimated the probability that an individual, with a specified combination of risk factors, would develop lung cancer within a 5-year period. Data from 579 lung cancer cases and 1157 age-and sex-matched population-based controls were available for this analysis. Significant risk factors were fitted into multivariate conditional logistic regression models. The final multivariate model was combined with agestandardised lung cancer incidence data to calculate absolute risk estimates. Combinations of lifestyle risk factors were modelled to create risk profiles. For example, a 77-year-old male non-smoker, with a family history of lung cancer (early onset) and occupational exposure to asbestos has an absolute risk of 3.17% (95% CI, 1.67 -5.95). Choosing a 2.5% cutoff to trigger increased surveillance, gave a sensitivity of 0.62 and specificity of 0.70, while a 6.0% cutoff gave a sensitivity of 0.34 and specificity of 0.90. A 10-fold cross validation produced an AUC statistic of 0.70, indicating good discrimination. If independent validation studies confirm these results, the LLP risk models' application as the first stage in an early detection strategy is a logical evolution in patient care. In addition, being the most common cancer with over 1.3 million incident cases per year, lung cancer has the highest worldwide rate of cancer mortality (Parkin et al, 2005). More than half of all cases are diagnosed at an advanced stage when surgical removal is no longer a viable treatment strategy. As a result, the overall 5-year survival rate is low, but stage-specific survival rates differ substantially by stage at presentation (van Rens et al, 2000). This raises the possibility that lung cancer may be an attractive candidate for screening, to detect disease at an early stage when treatment would be more effective. Recent results from the International Early Lung Cancer Action Program would appear to support this argument (I-ELCAP Investigators et al, 2006). While the International Early Lung Cancer Action Program results are very encouraging, there are also potential negative consequences of screening, including screen-detected false positives.Although a mortality benefit from spiral CT has not yet been confirmed in ongoing, large-scale randomised studies, the need to specify a high-risk target population is well accepted, and there has been increasing interest in methods of individual risk prediction for lung cancer. Models have been developed for use within highrisk groups (Bach et al, 2003), and for the general population (van Klaveren et al, 2002), although the latter tend to rely only on age and smoking. While epidemiological risk factors usually show poor discrimination between those who do and do not develop disease (Wald et al, 1999), lung cancer is an exception in that a high proportion of cases are attributable to one risk factor, smoking. However, there is room for further improvement in that many long-term smokers do not develop lung cancer. The predictive accuracy of lung cancer ri...