【作者】 林菁；

【导师】 何保昌；

【作者基本信息】 福建医科大学 ， 流行病与卫生统计学， 2021， 硕士

【摘要】 目的探讨口腔鳞状细胞癌(Oral squamous cell carcinoma,OSCC)患者淋巴结转移的影响因素,筛选和验证影响口腔鳞状细胞癌淋巴结转移的外泌体环状RNA(Circular RNAs,circ RNA)并构建术前淋巴结转移风险预测模型。方法1.回顾性分析506例OSCC患者颈部淋巴结转移的相关规律,并按3:1的比例将患者随机分成建模组(387例)和验证组(119例)。采用非条件Logistic回归模型评估计算建模组各影响因素与OSCC淋巴结转移风险关系的比值比(Odds ratio,OR)及其95%可信区间(Confidence interval,CI)。2.根据多因素分析结果筛选出建模组中口腔鳞状细胞癌淋巴结转移的影响因素,用于构建术前预测淋巴结转移风险的列线图(Nomogram)模型,并在建模组中评估模型预测性能以及使用验证组进行内部验证。3.采用circ RNA表达谱芯片扫描技术,在4例OSCC淋巴结转移患者与4例未转移患者术前血浆外泌体中筛选可能与淋巴结转移相关的circ RNA,在4对OSCC淋巴结转移患者术前与术后血浆外泌体中筛选可能来自于肿瘤的circ RNA,两者取交集后挑选在两者中同时呈显著差异表达的circ RNAs用于下一步的研究。4.采用实时荧光定量PCR(Quantitative Real-time PCR,q RT-PCR)方法在77例OSCC患者血浆外泌体中对筛选出的circ RNAs进行验证,并探讨circ RNAs的表达水平与OSCC淋巴结转移风险的关系。最后挑选与OSCC淋巴结转移有显著关系的circ RNA与对淋巴结转移有影响的临床特征构建Circ RNA-based nomogram模型。结果1.OSCC患者经病理学检查发现淋巴结转移率达到40.12%,颈部淋巴结转移主要发生在Ⅰ、Ⅱ、Ⅲ区,其中Ⅱ区淋巴结转移率最高(62.50%),而Ⅳ和Ⅴ区则较少累及。2.建模组多因素分析结果显示:活检分化程度为鳞状细胞癌较鳞状上皮病变的患者发生淋巴结转移的风险更大(OR=2.62,95%CI:1.38-4.98),颊癌及口底癌患者发生淋巴结转移的风险明显高于牙龈癌(OR=3.32,95%CI:1.20-9.20;OR=3.80,95%CI:1.13-12.80),MRI报告的淋巴结阳性的患者发生淋巴结转移的风险更大(OR=7.39,95%CI:3.55-15.40),此外,MRI报告的最大肿瘤直径≥2.6cm OSCC患者发生淋巴结转移的风险是<2.6cm患者的2.13倍(95%CI:1.26-3.60)。白细胞水平偏低与OSCC患者的淋巴结转移风险存在正相关关系(OR=5.46,95%CI:1.48-20.09)。3.基于建模组多因素分析结果筛选出的变量构建术前预测淋巴结转移风险的Nomogram模型,ROC和DCA结果显示Nomogram模型较MRI具有更好的预测性能以及更高的临床应用价值。4.芯片结果共发现7条肿瘤来源且与淋巴结转移有关的血浆外泌体circ RNA,其中上调2条,下调5条。通过circ RNA芯片表达丰度,引物设计效果,最终筛选出以下3个差异倍数最明显的外泌体hsa＿circ RNA＿047733、hsa＿circ RNA＿024144、hsa＿circ RNA＿403472用于下一步分析。5.q RT-PCR验证结果显示hsa＿circ RNA＿047733在OSCC淋巴结转移阳性组的表达水平低于淋巴结转移阴性组,表达水平较高的外泌体hsa＿circ RNA＿047733可以降低OSCC发生淋巴结转移的风险。6.Circ RNA-based nomogram模型在区分度、校准度、拟合优度、临床应用价值等方面均优于基于患者临床特征构建的Nomogram模型。结论1.活检分化程度、发病部位、MRI报告肿瘤最大径、MRI报告的淋巴结状态、白细胞水平是OSCC患者淋巴结转移的预测因子,基于以上5个指标构建的Nomogram模型对淋巴结转移风险预测性能良好。2.高表达的hsa＿circ RNA＿047733可以降低OSCC淋巴结发生转移的风险,在基于患者临床特征构建的Nomogram模型的基础上纳入血浆外泌体hsa＿circ RNA＿047733能够有效改善Nomogram模型对淋巴结转移风险的预测能力。更多还原

【Abstract】 ObjectiveTo explore the influencing factors of lymph node metastasis in patients with oral squamous cell carcinoma(OSCC),and to screen and verify the exosome circular RNA(circ RNA)related to lymph node metastasis of OSCC and to establish a model for predicting the risk of preoperative lymph node metastasis.Methods1.The distribution of cervical lymph node metastasis in 506 patients with OSCC was analyzed retrospectively.According to the proportion of 3:1,the patients were randomly divided into the modeling group(n=387)and the validation group(n=119).Unconditional logistic regression model was used to evaluate and calculate odds ratio(OR)and its 95% confidence interval(CI)between the influencing factors in the modeling group and the risk of lymph node metastasis in OSCC.2.According to the results of multivariate analysis,the influencing factors of lymph node metastasis of OSCC in the model group were selected to construct the nomogram model to predict the risk of lymph node metastasis before operation,and the prediction performance of nomogram model was evaluated in the modeling group,and the internal verification was carried out in the validation group.3.Four samples of preoperative plasma of OSCC with lymph node metastasis,4 samples of preoperative plasma of OSCC without lymph node metastasis and 4samples of postoperative plasma of OSCC with lymph node metastasis were selected for circ RNAs microarray detection.Then we demonstrated the difference of the expression of circ RNA between the patients of OSCC with lymph node metastasis before and after surgical treatment,OSCC with metastasis and without metastasis.The intersection of exosomal circ RNA of the three comparison group is used as the results of the initial screening.4.Circ RNAs was screened from plasma exocrine bodies of 77 patients with OSCC by real-time fluorescence quantitative PCR(q RT-PCR),and the relationship between the expression level of circ RNAs and the risk of lymph node metastasis of OSCC was studied.Finally,the circ RNA which was significantly related to the lymph node metastasis of OSCC and the clinical features that have influence on the lymph node metastasis are selected to construct the Circ RNA-based nomogram model.Results1.The overall lymph node metastasis rate of OSCC patients was 40.12%.Cervical lymph node metastasis mainly occurred in levels Ⅰ,Ⅱ and Ⅲ,of which levelⅡ had the highest lymph node metastasis rate(62.50%),while levels Ⅳ and Ⅴ were less involved.2.The results of multivariate analysis in the modeling group showed that patients with squamous cell carcinoma with differentiated degree of biopsy had a higher risk of lymph node metastasis than patients with squamous intraepithelial lesions(OR=2.62,95% CI: 1.38-4.98).The risk of lymph node metastasis in patients with buccal carcinoma and suboral carcinoma was significantly higher than that in gingival carcinoma(OR=3.32,95% CI: 1.20-9.20,OR=3.80,95% CI: 1.13-12.80),MRI reported that patients with positive lymph nodes had a higher risk of lymph node metastasis(OR=7.39,95% CI: 3.55-15.40).In addition,the risk of lymph node metastasis in patients with the largest tumor diameter ≥ 2.6cm OSCC reported by MRI was 2.13 times higher than that in patients with < 2.6cm(95% CI: 1.26-3.60).There was a positive correlation between low white blood cell level and the risk of lymph node metastasis in patients with OSCC(OR=5.46,95% CI: 1.48-20.09).3.Based on the variables selected from the multivariate analysis of the modeling group,a Nomogram model was constructed to predict the risk of preoperative lymph node metastasis.The results of ROC and DCA showed that Nomogram model had better predictive performance and higher clinical application value than MRI.4.The results of microarray showed that there were 7 plasma exosomal circ RNA from tumor tissue and related to lymph node metastasis,of which 2 were up-regulated and 5 down-regulated.Through the screening of the expression abundance of circ RNA microarray and the effect of primer design,it was found that the following three exosomes with the most obvious differences were used for further analysis: hsa＿circ RNA＿047733,hsa＿circ RNA＿024144 and hsa＿circ RNA＿403472.5.qRT-PCR showed that the expression level of hsa＿circ RNA＿047733 in OSCC lymph node metastasis positive group was lower than that in lymph node metastasis negative group.Hsa＿circ RNA＿047733 with higher expression level could reduce the risk of lymph node metastasis in OSCC.6.The Circ RNA-based nomogram model was superior to the Nomogram model based on patients’ clinical characteristics in terms of differentiation,calibration,goodness-of-fit and clinical application value.Conclusions1.The prediction factors of lymph node metastasis in patients with OSCC were the degree of differentiation of biopsy,the maximum diameter of tumor and the status of lymph nodes reported by MRI,tumor subsite and the level of white blood cells.The Nomogram model constructed in this study has a good performance in predicting the risk of lymph node metastasis.2.High expression of hsa＿circ RNA＿047733 could reduce the risk of lymph node metastasis in OSCC.The inclusion of exosomal hsa＿circ RNA＿047733 on the basis of Nomogram model based on patients’ clinical characteristics could effectively improve the ability of nomogram model to predict the risk of lymph node metastasis.更多还原