Lung cancer causes 1.8 million deaths worldwide and is the leading cause of cancer death globally. Although around 80% of lung cancer cases are attributed to tobacco smoking, in many Western countries a large proportion now occur in former or never smokers. Whilst a number of environmental and occupational exposures are known to increase the risk of lung cancer (for instance, radiation exposure), identifying additional factors could provide novel strategies for prevention.
However, investigating potential risk factors for lung cancer beyond smoking using traditional methods is particularly challenging, because many health conditions, socio-economic factors and molecular biomarkers strongly associate with smoking behaviours. An alternative approach, called Mendelian Randomisation (MR), can overcome this issue by estimating the association between genetically-predicted levels of an exposure and an outcome of interest. The method uses genetic variants that are randomly allocated and fixed before birth, which breaks their association with any external factors.
In a new study led by Dr Karl Smith-Byrne, Senior Molecular Epidemiologist in Oxford Population Health’s Cancer Epidemiology Unit, MR was used to screen hundreds of blood metabolites (chemicals formed during biochemical reactions that break down substances in the body) for an association with lung cancer risk. The results are published today in Cancer Epidemiology, Biomarkers & Prevention.
The researchers used a database of genetic variants (called single nucleotide polymorphisms, SNPs) that have previously been linked to 400 different metabolites in a genome-wide association study*. These SNPs were tested to see if any of them were also associated with the risk of lung cancer in a large genetic study of over 82,000 individuals. This resulted in three candidate metabolites with a potential link to lung cancer. After performing a series of rigorous tests (including to account for the fact that SNPs can by associated with multiple traits), only one metabolite remained that was robustly associated with lung cancer risk: isovalerylcarnitine (IVC).
The team then validated the MR finding using pre-diagnostic blood samples from a prospective case-control study of lung cancer, involving over 2,000 individuals (656 cases and 1,309 controls). This analysis confirmed that blood levels of IVC were inversely associated with lung cancer risk: each 10-fold increase in blood IVC was associated with 61% lower risk of lung cancer.
Dr Smith-Byrne said: ‘Our study is the first to integrate Mendelian Randomisation with traditional epidemiology methods as a novel approach to identify lung cancer biomarkers with a plausible role in causing this disease. Our study demonstrates an efficient way to triangulate plausible modifiable aetiological biomarkers that can readily be applied to other disease outcomes.’
Although it is currently unknown how IVC may affect lung cancer risk, according to the research team there are various potential mechanisms. IVC is a substrate of the enzyme isovaleryl-CoA dehydrogenase, which is involved in the degradation of the amino acid leucine and fatty acids. Leucine has a role in various intracellular signalling pathways that affect cell growth and proliferation, and also regulates the availability of glutamine, which may affect cancer proliferation and drug-resistance.
The researchers found no associations between the SNPs linked to IVC and smoking habits. Furthermore, the observational association between blood levels of IVC and lung cancer risk did not reduce after adjusting for smoking behaviours, indicating that the metabolite’s effect on lung cancer is likely to be independent of smoking.
Blood levels of IVC can be modified by a restricted protein diet or with glycine and L-carnitine supplementation, which may imply that dietary interventions could help reduce the risk of lung cancer for certain individuals.