Early Detection of Cancer Patients with Germline SDH Deficiency

初始训练集将包括40例SDH缺陷GIST患者，40例SDH缺陷副神经节瘤患者，以及40例罕见肿瘤自然史研究中性别和SDH亚基突变匹配的患者。外周血和尿样将在每天的第二个空收集，在三个不同的时间点收集，每个患者相隔至少一个月，以尽量减少正常变异性对代谢谱的影响。所有患者将继续使用目前的标准筛查建议进行随访。游离细胞DNA将从血浆中分离出来。尿cfDNA将使用Q-Sepharose色谱分离方案进行分离，该方案在经肾cfDNA产率方面优于其他已发表和商业化的尿cfDNA分离方法。PDL-1甲基化已被证明可以调节PDL-1的表达，DNA甲基化模式已被证明匹配或超过TMBs对PDL-1抑制的预测值。当分离到足够的cfDNA时，将使用先前发表的探针和数字PCR增强的Methylight技术评估PDL-1甲基化状态。该技术的最小DNA输入为3ng，检测限接近0.03%。PDL-1甲基化状态将与客观反应率进行比较，作为独立分析的预测价值，以及与cfDNA衍生的TMB结合时使用协同指数。虽然缺乏sdh的GIST基因组通常包含很少的体细胞突变，但已经观察到其他癌症相关基因，包括p53和RB，特别是在肿瘤行为更具侵袭性的患者中。除了甲基化分析，cfDNA变异等位基因检测和拷贝数改变也将使用CAPP-seq方法进行评估。先前验证和优化的探针序列将被纳入靶向混合捕获NGS面板（癌症个性化分析通过深度测序（CAPP-Seq））。利用集成数字误差抑制，罕见变异的CAPP-seq检测限接近0.0025%（2.5 / 105个分子）。代谢组学分析将与Naomi Taylor博士的实验室合作，通过Frederick NCI的质谱（蛋白质和小分子）核心进行。配对血清和尿液样本将在2小时内收集并冷冻。成批样品的代谢谱分析将通过hplc -质谱进行，包括TCA循环、戊糖磷酸分流和糖酵解代谢物的定量。目标2：模型构建将与NCI的癌症数据科学实验室合作进行。DNA甲基化和代谢组学数据的结合将为使用标准算法进行无监督聚类分析的模型开发提供强大的数据集。如果可能的话，模型将被细化和简化，只使用更好的区分特征。AIM 3：通过在罕见肿瘤自然史研究中持续招募生殖系SDH缺乏症患者，将生成一个测试数据集。代谢组学和cfDNA甲基化数据将从新入组的GIST、PHEO/PGL患者中收集，并且没有癌症证据（每组20例）。该模型将使用此数据集进行测试和改进。

The initial training set will include 40 patients with SDH-deficient GIST, 40 patients with SDH-deficient paraganglioma, and 40 patients matched for gender and SDH subunit mutation enrolled on the Rare Tumor Natural History Study. Peripheral blood and urine samples will be collected at the second void of the day at three separate time points separated by at least one month for each patient to minimize the impact of normal variability in the metabolic profiling. All patients will continue to be followed using the current standard screening recommendations. Cell free DNA will be isolated from plasma. Urine cfDNA will be isolated using a Q-Sepharose chromatography protocol which has been shown in the literature to outperform other published and commercial urine cfDNA isolation methods for trans-renal cfDNA yield. PDL-1 methylation has been demonstrated to regulate PDL-1 expression and DNA methylation patterns have been shown to match or exceed TMBs predictive value for PDL-1 inhibition. When sufficient cfDNA is isolated, PDL-1 methylation status will be assessed using previously published probes and digital PCR enhanced Methylight technology. This technology's minimal DNA input is 3ng with limits of detection approaching 0.03%. PDL-1 methylation status would be compared to objective response rates for predictive value as a stand-alone assay as well as when integrated with cfDNA derived TMB using a synergistic index. While the genome of SDH-deficient GIST typically contain very few somatic mutations, other cancer-associated genes including p53 and RB have been observed, particularly in patients with more aggressive tumor behavior. In addition to methylation analysis, cfDNA variant allele detection and copy number alterations will also be assessed using a CAPP-seq approach. Previously validated and optimized probe sequences will be incorporated into a targeted hybrid capture NGS panel (cancer personalized profiling by deep sequencing (CAPP-Seq)). Using integrated digital error suppression, CAPP-seq limit of detections for rare variants approaches 0.0025% (2.5 in 105 molecules). Metabolomic analysis will be performed in collaboration with the laboratory of Dr. Naomi Taylor through the Mass Spectrometry (Protein and Small Molecule) core at NCI at Frederick. Paired serum and urine samples will be collected and frozen within 2 hours of collection. Metabolic profiling of batched samples will be performed via HPLC-Mass spectrometry including quantitation of TCA cycle, pentose phosphate shunt, and glycolysis metabolites. AIM 2: Model building will be performed in collaboration with the NCI's Cancer Data Science Laboratory. The combination of DNA methylation and metabolomic data will provide a robust data set for model development using standard algorithms for unsupervised cluster analysis. If possible, the model will be refined and simplified to use only better-discriminating features. AIM 3: A testing data set will be generated through ongoing enrollment of patients with germline SDH deficiency on the Rare Tumor Natural History Study. Metabolomic and cfDNA methylation data will be collected from newly enrolled patients with GIST, PHEO/PGL, and without evidence of cancer (20 in each group). The model will be tested and refined using this dataset.