Validation of pancreatic cancer biomarkers in large prospective cohorts

在大型前瞻性队列中验证胰腺癌生物标志物

• 批准号: 8987610
• 负责人: ANNA E LOKSHIN
• 金额: $ 51.68万
• 依托单位: UNIVERSITY OF PITTSBURGH AT PITTSBURGH
• 依托单位国家: 美国
• 财政年份: 2015
• 资助国家: 美国
• 起止时间: 2015-05-05 至 2018-04-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/8987610
• 关键词: Acute; Algorithms; Benign; Bioinformatics; Biological Assay; Biological Markers; Cancer Etiology; Cessation of life; Classification; Clinical Trials; Colon; Data; Development; Diagnosis; Diagnostic; Disease; Early Diagnosis; Excision; Goals; Imaging Techniques; Individual; Inflammation; Lung; Malignant Neoplasms; Malignant neoplasm of pancreas; Operative Surgical Procedures; Ovarian; Pancreas; Pancreatic Ductal Adenocarcinoma; Patient Care; Patients; Performance; Phase; Probability; Procedures; Prostate; Proteomics; Resectable; Rewards; Risk; Sampling; Sensitivity and Specificity; Serum; Staging; Survival Rate; Symptoms; Time; United States; Validation; Women' s Health; base; case control; clinical practice; cohort; cost; differential expression; lymph nodes; mortality; outcome forecast; pre-clinical; prospective; public health relevance; resistin; screening; tumor

 DESCRIPTION (provided by applicant): Pancreatic ductal adenocarcinoma (PDAC) is the fourth leading cause of cancer death in the United States. The exceptionally poor prognosis of PDAC can be largely attributed to the difficulty of detecting the cancer at an early stage when curative surgical resection remains a viable option. The less than 10% patients who present with small, surgically resectable cancers have a realistic chance of cure and a 5-year survival rate of 20-30%. The identification of serum biomarkers associated with a high probability of PDAC could substantially change clinical practice by changing the risk-reward ratio for invasive imaging procedures or facilitating the development of less invasive procedures. At present, the overriding goal for the early detection of PDAC is the reliable identification of resectable tumors of less than 1.0 cm in size with negative lymph nodes. Presently, there are no biomarkers able to capture the disease more than 1 year prior to symptoms onset where there is a high probability of detecting PDAC at early stages. Biomarkers developed using samples obtained from patients at diagnosis do not validate well in preclinical samples indicating that different se of biomarkers characterizes early asymptomatic PDAC than late and symptomatic disease where biomarkers of inflammation and acute phase could predominate. We thus hypothesize that to generate a robust algorithm for identification of resectable tumors of less than 1.0 cm in size with negative lymph nodes, biomarker discovery should be performed in preclinical samples. We further hypothesize that biomarker velocity should be considered in order to generate optimized classification algorithm. We have generated a 5-biomarker classification algorithm that diagnoses patients 12-24 months before diagnosis with 743% SN at 98% SP and 24-35 months before diagnosis - with 64% SN at 98% SP. We have additionally accumulated promising preliminary data obtained in individual Women Health Initiative and pooled PLCO samples identifying biomarkers that are differentially expressed in preclinical cases vs. healthy controls and that demonstrate time-to-diagnoses dependent changes, thus providing support for our hypothesis that pre-clinical samples will be useful for discovery of pre-clinical biomarkers We propose to validate these candidate biomarkers in individual PLCO samples including longitudinal samples, develop an optimized classification algorithm, and validate this algorithm in preclinical samples from two other screening clinical trials. We propose following Specific Aims: 1. Optimize and validate the performance of preclinical PDAC classification algorithm in prospective sample. 2. Validate performance of a pre- diagnostic classification panel in two large prospective cohorts. 3. Construct Risk of Pancreatic Malignancy (RPM) algorithm to add the velocity component to PDAC classification. At the completion of the proposed study we expect to have identified a set of biomarkers with individual robust performance in preclinical PDAC samples, and have generated a bioinformatics algorithm based on biomarker velocities for recognition of PDAC at the resectable stages with high sensitivity and specificity.

 描述（由申请人提供）：胰腺导管腺癌（PDAC）是美国癌症死亡的第四大原因，PDAC 的预后极差，很大程度上归因于治愈性手术时早期发现癌症的难度。切除仍然是一种可行的选择，只有不到 10% 的可手术切除的小癌症患者有现实的治愈机会和 5 年生存率。 20-30% 与高概率 PDAC 相关的血清生物标志物的识别可以通过改变侵入性成像手术的风险回报比或促进微创手术的发展来显着改变临床实践。 PDAC 的早期检测是可切除肿瘤的可靠识别 目前，没有任何生物标志物能够在症状出现前 1 年内捕获该疾病，而在早期阶段检测到 PDAC 的可能性很高。诊断时的患者在临床前样本中没有得到很好的验证，这表明早期无症状 PDAC 的特征与晚期和有症状疾病不同，其中炎症和急性期的生物标记可能占主导地位，因此我们捕获了这一点以生成稳健的结果。对于识别小于 1.0 厘米且淋巴结阴性的可切除肿瘤的算法，应在临床前样本中进行生物标志物发现，我们进一步追求应考虑生物标志物速度，以生成优化的分类算法。生物标志物分类算法在诊断前 12-24 个月诊断患者，SN 为 98%，SP 为 743%；诊断前 24-35 个月，SN 为 98%，SP 为 64%。此外，还积累了在个体妇女健康倡议和汇总 PLCO 样本中获得的有希望的初步数据，这些数据识别出在临床前病例与健康对照中差异表达的生物标志物，并且证明了诊断时间依赖性变化，从而为我们的假设提供了支持，即临床前样本将有助于发现临床前生物标志物。我们建议在单个 PLCO 样本中纵向验证这些候选生物标志物，包括样本，开发优化的分类算法，并在来自其他两个临床筛选试验的临床前样本中验证该算法。我们提出以下具体目标： 1. 优化并验证临床前 PDAC 分类算法在前瞻性样本中的性能 2. 验证预诊断分类面板在两个大型前瞻性队列中的性能 3. 构建胰腺恶性肿瘤风险 (RPM) 算法。将速度分量添加到 PDAC 分类中。完成拟议的研究后，我们希望在临床前 PDAC 样本中识别出一组具有单独稳健性能的生物标志物，并生成一个基于生物标志物速度的生物信息学算法，用于在可切除阶段识别 PDAC，具有高灵敏度和特异性。