Proteomic, Glycomic and Autoantibody Lung Cancer Biomarker Validation

描述（由申请人提供）：肺癌是美国和世界范围内癌症死亡的主要原因。据估计，在2013年，美国将至少有228,000例新的肺癌确诊病例和超过159,000例死亡病例——大约相当于接下来四个最常见的癌症相关死亡原因的总和（结肠癌、乳腺癌、前列腺癌、胰腺癌）。肺癌早期检测的潜在好处是显而易见的——如果在早期发现，肺癌的5年生存率超过50%，而只有3.7%的转移性肺癌患者能存活这么长时间。迄今为止，大多数筛查方式（如胸部x线检查和痰细胞学检查）的随机对照试验尚未显示对肺癌死亡率有任何影响。此外，没有蛋白质组学或基因组血浆标记物在验证试验中取得足够的进展，可以作为fda批准的广泛筛选的可行候选物。在血液中发现可行的肺癌早期检测蛋白质组学、糖组学和/或免疫学生物标志物，这些标志物单独使用或与低剂量胸部计算机断层扫描（CT）结合使用将特别有价值，因为肺癌是常见的，通常是致命的，治疗费用昂贵-因此它具有很高的社会成本。我们采取了一种独特的方法来发现生物标志物，旨在克服当前的障碍。我们创建了一个高密度抗体阵列，包含3200种不同的抗体，我们使用它来询问来自多个观察性试验的预诊断样本集（即，数百个具有良好匹配的病例和对照的预诊断样本），以评估蛋白质组学，糖糖和自身抗体的差异。由于我们的发现方法使用高亲和力抗体，我们最好的标记可以很容易地转移到elisa样方法，促进正式验证和进入临床。我们已经证明该技术具有高灵敏度（低图形水平）和可重复性（大多数变异系数<10%）。此外，我们已经确认并发现了卵巢癌、乳腺癌、结肠癌和肺癌中新的可行的蛋白质组学生物标志物候选物。使用心血管健康研究（CHS）的诊断前肺癌样本，我们发现30个蛋白质组学、糖糖或自身抗体生物标志物在诊断为肺癌的人群中显著增加（p<0.002）。在这里，我们建议使用来自前列腺、肺、结直肠和卵巢（PLCO）癌症筛查试验的样本来验证这些候选药物。我们的具体目标是：(1)利用PLCO样品初步验证30个潜在的肺癌早期检测蛋白质组学、糖组学或自身抗体生物标志物候选物（例如，>增加2倍，p<0.002, AUC > 0.67）。(2)将Aim 1中评估的生物标志物与多种临床和流行病学因素（包括肺癌亚型、吸烟史、年龄、性别、体重指数、肺癌和其他癌症家族史、辐射、石棉和其他暴露）结合起来，开发和评估一种新型风险预测模型的性能，这些因素可能会影响生物标志物的性能和风险预测。

DESCRIPTION (provided by applicant): Lung cancer is the leading cause of cancer death in the United States and world-wide. In 2013, it is estimated that there will be at least 228,000 new cases of lung cancer diagnosed and more than 159,000 deaths in the United States - approximately equal to the next four most common causes of cancer-related mortality combined (colon, breast, prostate, pancreas). The potential benefits of lung cancer early detection are clear - when discovered at an early stage, lung cancer has 5 year survival rates over 50% while only 3.7% of metastatic lung cancer patients survive that long. To date, randomized controlled trials of most screening modalities such as chest x-ray and sputum cytology have not demonstrated any impact on lung cancer mortality. Also, no proteomic or genomic plasma markers have advanced sufficiently in validation trials to be viable FDA-approved candidates for widespread screening. Discovery of viable lung cancer early detection proteomic, glycomic and/or immunological biomarkers in blood that would work well on their own or in combination with low-dose chest computed tomography (CT) would be especially valuable since lung cancer is common, most often fatal and is expensive to treat - so it has high societal costs. We have taken a unique approach to biomarker discovery designed to overcome current obstacles. We created a high density antibody array containing 3200 different antibodies that we use to interrogate pre-diagnostic sample sets from multiple observational trials (i.e., hundreds of pre-diagnostic samples with well-matched cases and controls) in a nested case-control design study to evaluate proteomic, glycomic and autoantibody differences. Since our discovery method uses high affinity antibodies, our best performing markers can be readily transferred to ELISA-like methods facilitating formal validation and movement into the clinic. We have shown the technology is highly sensitive (low picogram levels) and reproducible (most coefficients of variation <10%). Furthermore, we have confirmed known and found new viable proteomic biomarker candidates in ovarian, breast, colon and lung cancer. Using pre-diagnostic lung cancer samples from the Cardiovascular Health Study (CHS), we found 30 proteomic, glycomic or autoantibody biomarkers that were significantly increased (p<0.002) in people that are diagnosed with lung cancer. Here, we propose to use samples from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial to validate these candidates. Our specific aims are: (1) To use PLCO samples to preliminarily validate 30 potential lung cancer early detection proteomic, glycomic or autoantibody biomarker candidates (e.g., those with >2 x increases, p<0.002, AUC > 0.67). (2) To develop and assess the performance of a novel risk prediction model integrating the biomarkers assessed in Aim 1 with a variety of clinical and epidemiologic factors, including lung cancer subtype, smoking history, age, gender, body mass index, family history of lung and other cancers, radiation, asbestos and other exposures, that should potentially impact biomarker performance and risk prediction.