The International Lung Cancer Consortium (ILCCO) was established in 2004 to maximize research efficiency for lung cancer and to share comparable epidemiological and clinical data, and biological samples across studies. Since its establishment, over 70 studies have participated in the ILCCO and shared comparable clinico-epidemiological data and a subset with biological samples and genomic data. The data harmonization was conducted at the Sinai Health System in Toronto, and genomic data is managed at the Dartmouth College of Medicine/Baylor College of Medicine. In total, the ILCCO Data Repository now has epidemiological data for over 1.2 million study participants, including 100,000 lung cancer patients, and genomic data on approximately 50,000 study participants. The large-scale epidemiological and genomic data allow us to extensively study and characterize the etiological factors, including lifestyle risk factors, medical history and genomic architectures for lung cancer development. Data submitted from all studies are systematically checked for missing values, outliers, inadmissible values, aberrant distributions and internal inconsistencies before harmonization. Common variable definitions were developed. For lifestyle risk factors and medical history, we conducted meta-analysis based on study-specific estimates, when applicable. If heterogeneities were present, random effects models were employed to account for the heterogeneity across studies. For subgroup of interests or when sample size is limited, pooled-analyses based on individual-level data were applied. When applicable, the non-linearity relationship was assessed. For genetic susceptibility of lung cancer, we investigated the genetic loci associated with lung cancer risk using log-additive model adjusted for population ancestry and account for multiple comparisons. To assess the causality of specific exposures and lung cancer risk, we applied Mendelian Randomization and mediation analytical approaches. To estimate 5-year lung cancer absolute risk, we incorporated risk factors, medical history and genetic factors based on age-specific lung cancer incidence and the competing risk. Based on 17 ILCCO studies (24,000 cases and 81,000 controls), we observed a robust association between lung cancer risk and emphysema and pneumonia, even among never smokers, and after long latency period. Based on 24 ILCCO studies, we quantified the association between family history of lung cancer and its risk by their smoking status and affected relative types. Based on 6 studies in UK, Canada, UK and New Zealand, we assessed the association between cannabis smoking and lung cancer risk by intensity, duration and cumulative exposures and by histological subtypes. We have recently completed a largest lung cancer genetic analysis based over 29,000 lung cancer cases and 56,000 controls. We identified 10 novel lung cancer susceptibility loci, in addition to the known regions, such as TERT/CLPTM1L, CHRNA5, MHC region, RAD52, CHEK2 and found specific associations mediated through mRNA expression. We helped to quantify the effect of specific genetic variant in nicotinic receptor gene on smoking cessation and age of onset. Using genetic instruments and Mendelian Randomization approach, we confirmed the association between lung cancer risk and long telomere length. Most recently, we investigated the association between impaired lung function and lung cancer risk based on UK Biobank and ILCCO OncoArray data, and we found that impaired lung function was associated with lung cancer risk in never smokers and particularly for adenocarcinoma, most likely through immune-mediated pathways. When combining all factors into an integrative risk model, we found that individuals with highly polygenic risk scores reached lung cancer screening threshold at younger age than those with average genetic risk background. ILCCO provides a powerful research platform for research on lung cancer. The collaborative projects based on ILCCO have contributed to the understanding of lung cancer etiology beyond tobacco smoking. As future perspectives, ILCCO has obtained clinical prognosis data for over 50,000 lung cancer patients and will also be able to investigate factors associated with lung cancer prognosis in depth. Finally, ILCCO has built close collaborations with several lung cancer low-dose computed tomography screening programs to jointly investigate the optimal strategy for risk stratification and early detection for lung cancer.
Read full abstract