Evaluation of Pain in Chronic Pancreatitis using the NAPS2 cohorts

描述（由申请人提供）：慢性胰腺炎（CP）是一种进行性和破坏性胰腺炎性疾病。最令人痛苦的特征之一是疼痛，约90%的患者发生疼痛，其中约一半患者持续疼痛，这与住院时间增加和生活质量降低有关。由于存在多种疼痛病因，并且由于没有标准化的方法来评估CP患者的疼痛，因此需要新的方法来更好地了解CP疼痛的机制以及如何进行适当的治疗。北美胰腺炎研究2（NAPS 2）从美国最大的胰腺炎队列中招募、表型分析并获得生物标本，我们已经完成了一项全基因组关联研究（GWAS），其中包括1，171例NAPS 2 CP病例，这些病例的详细表型疼痛数据可用。我们建议分析NAPS 2数据集和生物标本（DNA，血清，胰腺组织），以确定持续疼痛的遗传风险和潜在机制，识别炎性疼痛的标志物，并表征临床疼痛复合物，以指导患者管理。我们已经通过完整的和嵌套的GWAS确定了8个“持续”疼痛的候选基因。目的1将确定这些变异是否与恒定疼痛表型相关基因的功能变化有关。将进行靶向SNP基因分型和基因表达研究，包括人类样本中的mRNA研究和免疫组织化学染色。目的2将确定是否c-反应蛋白（CRP）或细胞因子生物标志物与持续疼痛相关。疼痛是活动性炎症的指标，促炎细胞因子增加疼痛，而抗炎细胞因子减少疼痛。我们将测试500个NAPS 2样本的CRP升高和10种Th 1/Th 2细胞因子作为炎症的生物标志物。对于阳性对照，我们将包括来自40名急性胰腺炎患者的血液样品，这些患者因疼痛或炎症而住院> 4天。目标3将应用机器学习方法来测试基因型、生物标志物和形态学（梗阻）与疼痛模式、严重程度和特征的定量测量以及SF 12 v2生活质量评分（精神和身体）的相关性，同时控制 性，吸烟，酗酒我们预计，我们的机器学习方法将根据值得正式测试的病因学为疼痛的正确分类提供决策规则。 在临床试验阶段此外，预计机器学习的使用将通过将症状特征、生物标志物、成像研究和遗传学与疼痛性CP患者中观察到的复杂机制最佳地联系起来，从而提供对派机制的深入了解。这项研究的目标是利用现有的NAPS 2数据和生物标本为未来的临床研究构建一个框架，以测试基于我们的机器学习预测病因而不是症状的个性化疼痛管理的有效性。

DESCRIPTION (provided by applicant): Chronic pancreatitis (CP) is a progressive and destructive inflammatory disorder of the pancreas. One of the most distressing features is pain, which occurs in ~90% of patients, about half of whom have constant pain, which is associated with increased hospitalization and lower quality of life. Since there are multiple pain etiologies and since no standardized method exists to assess pain in patients with CP, novel approaches are needed to better understand the mechanisms of pain in CP and how it can be appropriately treated. The North American Pancreatitis Study 2 (NAPS2) recruited, phenotyped, and obtained biospecimens from the largest US cohort of pancreatitis, and we have completed a genome-wide association study (GWAS) that includes 1,171 NAPS2 CP cases for whom detailed phenotypic pain data are available. We propose to analyze the NAPS2 data set and biospecimens (DNA, serum, pancreas tissue) to define genetic risks and potential mechanisms of constant pain, identify markers of inflammatory pain, and characterize clinical pain complexes that can guide patient management. We have already identified through the full and a nested GWAS 8 candidate genes for "constant" pain. Aim 1 will determine whether these variants are associated with functional changes in genes associated with the constant pain phenotype. Targeted SNP genotyping and gene expression studies will be conducted, including mRNA studies and immunohistochemical staining in human samples. Aim 2 will determine whether c-reactive protein (CRP) or cytokine biomarkers correlate with constant pain. Pain is an indicator of active inflammation, with pro-inflammatory cytokines increasing pain, and anti-inflammatory cytokines diminishing pain. We will test 500 NAPS2 samples for elevated CRP and 10 Th1/Th2 cytokines as biomarkers of inflammation. For positive controls we will include blood samples from 40 acute pancreatitis patients who remain hospitalized for > 4 days for pain or inflammation. Aim 3 will apply machine-learning approaches to test for correlation of genotype, biomarkers, and morphology (obstruction) with quantitative measures of pain pattern, severity, and character as well as SF12 v2 quality of life scores (mental and physical) while controlling for sex, smoking, and alcohol. We anticipate that our machine learning approaches will provide decision rules for the proper classification of pain according to etiology worthy of formal testing in clinical trials. In addition, use of machine learning is anticipated to provide insight into pai mechanism by optimally linking the symptoms signatures, biomarkers, imaging studies, and genetics to complex mechanisms that are seen in patients with painful CP. The goal of this study is to use existing NAPS2 data and biospecimens to construct a framework for future clinical studies that test the effectiveness of personalized pain management based on our machine-learning-predicted etiology rather than symptoms alone.