A Propensity Score-adjusted Analysis of the Effects of Ubiquitin E3 Ligase Copy Number Variation in Peripheral Blood Leukocytes on Colorectal Cancer Risk
A Propensity Score-adjusted Analysis of the Effects of Ubiquitin E3 Ligase Copy Number Variation in Peripheral Blood Leukocytes on Colorectal Cancer Risk
复制标题
DOI：
10.7150/jca.29872
复制
发表时间：
2019-01-01
影响因子：
3.9
通讯作者：
Zhao, Yashuang
中科院分区：
文献类型：
--
作者：
Bi, Haoran;Liu, Yupeng;Zhao, Yashuang
背景：泛素连接酶E3（E3 s）在许多致癌生物过程中的特异性蛋白质降解中起关键作用。大肠癌（CRC）的发展可能受到E3基因拷贝数变异（CNV）的影响。以往的研究可能低估了潜在的混杂因素对基因CNV和CRC风险之间的关联的影响，并且缺乏整合基因CNV模式的CRC风险预测模型。我们的研究试图通过使用倾向评分（PS）调整和开发整合CNV模式用于CRC风险预测的模型来评估MDM 2、SKP 2、FBXW 7、β-TRCP和NEDD 4 -1的基因CNV和CRC风险。这项研究包括1036名参与者，使用传统回归和不同的PS技术来调整混杂因素，以评估五个基因CNV与CRC风险之间的关系，建立CRC风险预测模型。应用AUC评价模型的效果。结果：与变量校正相比，PS校正后的比值比（OR）趋于保守和准确，置信区间（CI）较窄。PS调整后，MDM 2扩增与CRC风险增加相关（Amp模式：OR = 8.684，95%CI：1.213-62.155，P = 0.031），而SKP 2缺失和（del+amp）基因型与CRC风险降低相关（Del模式：OR = 0.323，95% CI：0.106-0.979，P = 0.046; Var-pattern：OR = 0.339，95% CI：0.135-0.854，P = 0.024）。整合基因CNV模式的预测模型可以正确地重新分类1.7%的subjects.Conclusions：MDM 2扩增和SKP 2 CNV分别与增加和降低CRC风险相关;异常CNV整合模型预测CRC风险更精确。需要进一步的研究来验证这些令人鼓舞的结果。
Background: The ubiquitin ligases E3 (E3s) plays a key role in the specific protein degradation in many carcinogenic biological processes. Colorectal cancer (CRC) development may be affected by the copy number variation (CNV) of E3s. Prior studies may have underestimated the impact of potential confounding factors' effects on the association between gene CNV and CRC risk, and CRC risk predictive model integrating gene CNV patterns is lacking. Our research sought to assess the genes CNVs of MDM2, SKP2, FBXW7, beta-TRCP, and NEDD4-1 and CRC risk by using propensity score (PS) adjustment and developing models that integrate CNV patterns for CRC risk predictions.Methods: This study comprising 1036 participants used traditional regression and different PS techniques to adjust the confounding factors to evaluate the relationships between five gene CNVs and CRC risk, and to establish a CRC risk predictive model. The AUC was applied to evaluate the effect of the model. The categorical net reclassification improvement (NRI) and the integrated discrimination improvement (IDI) were analyzed to evaluate the discriminatory accuracy improvement among the models.Results: Compared to variable adjustment, the odds ratios (ORs) tended to be conservative and accurate with narrow confidence intervals (CIs) after PS adjustment. After PS adjustment, MDM2 amplification was related to increased CRC risk (Amp-pattern: OR = 8.684, 95% CI: 1.213-62.155, P = 0.031), whereas SKP2 deletion and the (del+amp) genotype were associated with reduced CRC risk (Del-pattern: OR = 0.323, 95% CI: 0.106-0.979, P = 0.046; Var-pattern: OR = 0.339, 95% CI: 0.135-0.854, P = 0.024). The predictive model integrating the gene CNV pattern could correctly reclassify 1.7% of the subjects.Conclusions: MDM2 amplification and SKP2 CNVs are associated with increased and decreased CRC risk, respectively; abnormal CNV-integrated model is more precise for predicting CRC risk. Further studies are needed to verify these encouraging outcomes.