Tricia L Larose, Arnulf Langhammer and Mattias Johansson, for the Lung Cancer Cohort Consortium (LC3)
Lung cancer is one of the most common cancers worldwide, accounting for 2.09 million cases and 1.76 million deaths in 2018. Two of the most prolific cancer epidemiologists of our time — Sir Richard Doll and Sir Bradford Hill — identified smoking as the biggest cause of lung cancer in their seminal report, “Smoking and Carcinoma of the Lung”, published in the British Medical Journal in 1950. Nearly 70 years later, smoking remains the predominant risk factor for lung cancer, as well as for 15 additional cancers and other non-communicable diseases.
Cigarettes contain nicotine — a highly addictive substance that induces pleasure and reduces stress and anxiety. Many current smokers would like to quit but, due to nicotine addiction, more than 80% of those who attempt to quit return to smoking within 6 months. Moreover, smoking may be a social norm in some familial or social clusters, thus reinforcing the behaviour despite the well-known negative health effects.
In our recent study published in the International Journal of Epidemiology, we studied the lung cancer risk implications of blood concentrations of cotinine — a nicotine metabolite and an objective measure of recent tobacco exposure. In particular, we measured pre-diagnostic cotinine concentrations for nearly 5500 people who were later diagnosed with lung cancer and 5500 additional matched control participants from 20 prospective cohort studies in the US, Europe, Asia and Australia. This is by far the largest study of its kind to date.
The median time between blood sampling and lung cancer diagnosis in the study participants was 6.3 years. We considered cotinine concentrations of 115 nmol/L or higher to be indicative of active smoking, between 5 and 115 nmol/L as indicative of second-hand smoke exposure, and less than 5 nmol/L as unexposed to smoking.
Our findings were threefold. First, we showed that circulating cotinine concentrations are consistently associated with lung cancer risk for current smokers, over and above that indicated by self-reported smoking exposure. Second, we showed that combining cotinine measures with self-reported smoking may help identify individuals at elevated risk of developing lung cancer, compared with relying on self-reported smoking information alone. Finally, when we compared self-reported smoking status (never, former or current smokers) with circulating cotinine concentrations, we found cotinine concentrations consistent with active smoking to be common in former smokers (cases: 14.6%, controls: 9.2%).
To evaluate the association between exposure to second-hand smoke and lung cancer risk, we estimated odds ratios for former and never smokers separately by comparing participants with cotinine concentrations between 5 and 115 nmol/L (exposed) to participants with cotinine concentrations below 5 nmol/L (unexposed). However, despite our relatively large sample of more than 1500 former and 1300 never smoker case–control pairs, we did not observe a risk increase for participants with circulating cotinine concentrations consistent with second-hand smoke exposure.
Our results highlight that misclassification of self-reported smoking status may be common in epidemiological studies. This is an important consideration for epidemiological studies on smoking-related diseases, as misclassification among self-reported former smokers may result in a failure to fully account for smoking as an underlying reason for an observed association between a putative risk factor and disease risk — a phenomenon often referred to as “residual confounding” by epidemiologists. Considering the impact of tobacco exposure on the risk of a wide range of diseases, as well as its influence on most risk factors, it is virtually impossible to exclude tobacco exposure as an underlying reason for many alleged risk factor–disease relationships. To this end, cotinine — as an objective measure of recent smoking intensity — offers a means to circumvent this inherent limitation of many epidemiological studies.
The findings of our study also have important clinical implications. Risk prediction models that combine pre-diagnostic biomarker measures with other self-reported data may better identify patients who would benefit from lung cancer screening (e.g. CT screening). A recent large clinical trial has proven that early detection of lung cancer through screening can decrease mortality rates by over 20%. However, only about 50% of people who are diagnosed with lung cancer are eligible for screening using current recommended criteria. With biomarker-based risk prediction models, it may be possible to improve screening effectiveness by better identifying those individuals who are most likely to benefit from screening. Our study demonstrated that cotinine, perhaps in combination with other risk-indicative biomarkers, may be of use for that purpose.
Read more:
Larose TL, Guida F, Fanidi A, et al. Circulating cotinine concentrations and lung cancer risk in the Lung Cancer Cohort Consortium (LC3). Int J Epidemiol 2018; 47: 1760-1771. https://doi.org/10.1093/ije/dyy100.
Tricia L Larose, PhDis a postdoctoral scientist with the Genetic Epidemiology Group, Section of Genetics, International Agency for Research on Cancer — the specialised cancer agency of the World Health Organization. She is cross-appointed to the K.G. Jebsen Center for Genetic Epidemiology, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology. Her work is supported by the Research Council of Norway (grant number 267776/H10). You can follow Tricia on Twitter @TricLarose
Arnulf Langhammer, MD, PhD is a medical doctor and clinician, a professor and leader of the HUNT databank. The HUNT study is one of 20 participating cohorts in the Lung Cancer Cohort Consortium (LC3).
Mattias Johansson, PhD is a scientist in the Genetic Epidemiology Group, Section of Genetics, International Agency for Research on Cancer, World Health Organization. He is the co-PI for the Lung Cancer Cohort Consortium (LC3). The LC3 was supported by the National Institutes of Health/National Cancer Institute (grant number 1U1CA155340-01) and the Australian National Health and Medical Research Council (grant number 1050198). You can follow Mattias on Twitter @Mattias31
IJEblog 