Covariate-based linkage analysis of lung cancer risk reveals novel loci on 9p21 and 20q12

Tuesday, October 25, 2011 — Poster Session II
Noon – 2:00 p.m.	Natcher Conference Center	NHGRI	GEN/GENOM-16

Authors

• C Simpson
• T Green
• B Doan
• C Amos
• S Pinney
• E Kupert
• M de Andrade
• P Yang
• A Schwartz
• P Fain
• A Gazdar
• J Minna
• J Wiest
• H Rothschild
• D Mandal
• M You
• T Coons
• C Gaba
• M Anderson
• J Bailey-Wilson

Abstract

Lung cancer (LC) is a major killer in the US (over 150,000 deaths in 2010). Environmental risk factors like smoking and asbestos exposures are well known. However, only 15% of smokers develop LC, suggesting genetic effects or gene-environment (GxE) interactions. We previously mapped a major LC susceptibility locus to 6q23-q25, with a rare risk haplotype in linked families that exhibit GxE interaction between the locus and smoking. Genome-wide association studies (GWAS) suggest other loci with common alleles of small effect on LC risk. However, these loci do not explain all familial risk of LC, suggesting that additional risk alleles exist. We used LODPAL from SAGE to perform linkage analysis contrasting identity-by-descent sharing in affected and discordant relative pairs, adjusting for a single environmental covariate (propensity score for LC risk based on pack-years of smoking and its square, age and sex). Using only one covariate increases power in LODPAL since each covariate adds a degree of freedom to the test. Strong evidence of linkage to LC was observed on 6p (LOD=5.72, 74cM) and 6q (LOD=3.25, 173cM) and 20q12 (LOD=3.42, 63cM). Linkage to lung/throat cancer was observed on 9p21 (LOD=5.66, 64cM). Permutations are ongoing to determine empirical p-values for these LOD scores.

back to top