Young Robert P, Scott Raewyn J, Callender Tom, Duan Fenghai, Billings Paul, Aberle Denise R, Gamble Greg D
Faculty of Medical and Health Sciences, University of Auckland, Auckland P.O. Box 37-971, New Zealand.
Respiratory Research Group, Greenlane Clinical Centre, Epsom, Auckland 1344, New Zealand.
J Clin Med. 2025 Apr 30;14(9):3110. doi: 10.3390/jcm14093110.
Epidemiological studies suggest lung cancer results from the combined effects of smoking and genetic susceptibility. The clinical application of polygenic risk scores (PRSs), derived from combining the results from multiple germline genetic variants, have not yet been explored in a lung cancer screening cohort. This was a post hoc analysis of 9191 non-Hispanic white subjects from the National Lung Screening Trial (NLST), a sub-study of high-risk smokers randomised to annual computed tomography (CT) or chest X-ray (CXR) and followed for 6.4 years (mean). This study's primary aim was to examine the relationship between a composite polygenic risk score (PRS) calculated from 12 validated risk genotypes and developing or dying from lung cancer during screening. Validation was undertaken in the UK Biobank of unscreened ever-smokers (N = 167,796) followed for 10 years (median). In this prospective study, we found our PRS correlated with lung cancer incidence ( < 0.0001) and mortality ( = 0.004). In an adjusted multivariable logistic regression analysis, PRS was independently associated with lung cancer death ( = 0.0027). Screening participants with intermediate and high PRS scores had a higher lung cancer mortality, relative to those with a low PRS score (rate ratios = 1.73 (95%CI 1.14-2.64, = 0.010) and 1.89 (95%CI 1.28-2.78, = 0.009), respectively). This was despite comparable baseline demographics (including lung function) and comparable lung cancer characteristics. The PRS's association with lung cancer mortality was validated in an unscreened cohort from the UK Biobank ( = 0.002). In this biomarker-based cohort study, an elevated PRS was independently associated with dying from lung cancer in both screening and non-screening cohorts.
流行病学研究表明，肺癌是吸烟和遗传易感性共同作用的结果。多基因风险评分（PRS）是通过综合多个种系基因变异的结果得出的，其在肺癌筛查队列中的临床应用尚未得到探索。这是一项对来自国家肺癌筛查试验（NLST）的9191名非西班牙裔白人受试者进行的事后分析，该试验是一项针对高危吸烟者的子研究，这些吸烟者被随机分配接受年度计算机断层扫描（CT）或胸部X线（CXR）检查，并随访6.4年（平均）。本研究的主要目的是检验由12种经过验证的风险基因型计算得出的复合多基因风险评分（PRS）与筛查期间肺癌发生或死亡之间的关系。在英国生物银行中对未接受筛查的既往吸烟者（N = 167,796）进行了验证，随访10年（中位数）。在这项前瞻性研究中，我们发现我们的PRS与肺癌发病率（<0.0001）和死亡率（= 0.004）相关。在调整后的多变量逻辑回归分析中，PRS与肺癌死亡独立相关（= 0.0027）。与低PRS评分者相比，PRS评分中等和高的筛查参与者肺癌死亡率更高（率比分别为1.73（95%CI 1.14 - 2.64，= 0.010）和1.89（95%CI 1.28 - 2.78，= 0.009））。尽管基线人口统计学特征（包括肺功能）和肺癌特征相当。PRS与肺癌死亡率的关联在英国生物银行的未筛查队列中得到验证（= 0.002）。在这项基于生物标志物的队列研究中，升高的PRS在筛查和未筛查队列中均与死于肺癌独立相关。
