【作者】 刘菲；

【导师】 杨兴升；

【作者基本信息】 山东大学 ， 妇产科学（专业学位）， 2024， 硕士

【摘要】 目的:子宫内膜癌是女性最常见的癌症之一,分期为Ⅰ期的患者预后较好,5年生存率约为95%,但若出现区域淋巴结转移,生存率则降低至69.4%。淋巴脉管间隙浸润(Lymph-vascular space invasion,LVSI)与局部淋巴结转移密切相关,被认为是区域淋巴结转移的先决条件。术前明确是否有LVSI在Ⅰ期子宫内膜癌患者是否需行淋巴结清扫术方面具有参考价值。关于LVSI是否会对Ⅰ期子宫内膜癌的预后产生不良影响,不同的研究持有不同的观点。本研究旨在分析Ⅰ期子宫内膜癌发生LVSI的高危因素,建立Ⅰ期子宫内膜癌LVSI风险的列线图预测模型,并评估LVSI对Ⅰ期EC患者预后的影响。方法:本研究收集2014年1月至2019年12月就诊于山东大学齐鲁医院并行全面分期手术的749例Ⅰ期EC患者的病历资料。将纳入研究的749例患者按照7:3的比例使用SPSS 27.0软件分配至训练组或者验证组。按照术后常规病理报告的LVSI状态,将患者分为淋巴脉管间隙浸润阴性组和阳性组,对训练组患者的基本特征进行组间相关性分析,并运用多因素分析确定Ⅰ期EC患者发生淋巴脉管间隙浸润的独立危险因素。运用R语言软件(R4.2.3)在训练组中构建Ⅰ期子宫内膜癌LVSI风险的列线图预测模型。采用ROC曲线、校准曲线及DCA曲线分别在训练组及验证组中对构建的模型进行内部验证及外部验证以评估模型性能。本研究的预后结局事件设定为复发或者死亡,评价指标设定为无病生存期(disease-freesurvival,DFS)或总生存期(overallsurvival,OS)。利用单因素及多因素Cox比例风险回归分析探讨Ⅰ期EC患者预后的独立危险因素。采用倾向匹配评分算法(propensity score matching,PSM)平衡纳入研究的Ⅰ期EC患者LVSI阴性组及LVSI阳性组患者的临床基本特征,并利用Kaplan-Meier生存分析探究LVSI对Ⅰ期子宫内膜癌预后的影响。结果:1.Ⅰ期子宫内膜癌LVSI风险预测模型的建立及验证本研究共纳入749例行全面分期手术的Ⅰ期子宫内膜癌患者,其中LVSI阴性患者608例,LVSI阳性患者141例。将所有患者按7:3比例随机分配至训练组和验证组。其中训练组528例,验证组221例。训练组的组间相关性分析及Logistic多因素回归分析结果显示:高血压、CA-125水平、肿瘤最大直径、病理类型、肌层浸润深度、组织学分级与Ⅰ期子宫内膜癌发生LVSI显著相关(P<0.05);CA-125>35 U/ml(OR=2.420,95%CI:1.369～4.276,P=0.002),肿瘤最大直径>2cm(OR=2.032,95%CI:1.214～3.401,P=0.007),病理类型为非子宫内膜样癌(OR=4.792,95%CI:2.249～10.209,P<0.001),组织学分级为 G3(OR=3.449,95%CI:1.891～6.291,P<0.001),肌层浸润深度≥1/2(OR=2.408,95%CI:1.354～4.284,P=0.003)是Ⅰ期子宫内膜癌发生LVSI的独立危险因素。利用筛选出的CA-125水平、肿瘤最大直径、病理类型、组织学分级及肌层浸润深度五个独立危险因素作为预测因子构建Ⅰ期子宫内膜癌LVSI风险预测模型。采用ROC曲线、校准曲线及DCA曲线分别在训练组及验证组中对该模型进行内部验证及外部验证,结果显示该模型具有良好的准确性、校准度及临床效能。2.LVSI对Ⅰ期子宫内膜癌预后影响及预后独立危险因素本研究随访时间为22-117个月,在随访期间,有19(2.5%)例患者因本病死亡,34(4.5%)例患者疾病复发。5年总生存率为95.1%,5年无疾病复发生存率为93.2%。在纳入研究的所有Ⅰ期EC患者中,单因素及多因素Cox比例风险回归分析结果显示:组织学分级为 G3(HR=3.441,95%CI:1.296-9.134,P=0.013),LVSI 阳性(HR=3.503,95%CI:1.242-9.883,P=0.018)是Ⅰ期子宫内膜癌DFS的独立危险因素;组织学分级为G3(HR=6.404,95%CI:1.798-22.817,P=0.004),肌层浸润深度≥1/2(HR=4.036,95%CI:1.101-14.788,P=0.035),LVSI 阳性(HR=5.151,95%CI:1.346-19.710,P=0.017)是Ⅰ期子宫内膜癌患者OS的独立危险因素。采用PSM对LVSI阴性及LVSI阳性两组患者的组间特征进行平衡计算。在匹配前队列中,LVSI阴性组和LVSI阳性组间的无病进展期(P<0.001)及总生存期(P=0.002)的差异均具有统计学意义;在匹配后队列中,两组间DFS(P=0.037)及OS(P=0.043)的差异仍均具有统计学意义。结论:1.CA-125>35U/ml,肿瘤最大直径>2cm,病理类型为非子宫内膜样癌,高级别病变,肌层浸润深度≥1/2是Ⅰ期EC患者发生淋巴脉管间隙浸润的独立危险因素。2.本研究所建立的Ⅰ期EC患者LVSI风险预测模型具有良好的准确性、校准度及临床效能,可帮助临床医生术前初步预估Ⅰ期子宫内膜癌患者LVSI阳性的风险;3.LVSI 阳性是Ⅰ期EC预后不良的独立危险因素,对DFS及OS均会产生不良影响。更多还原

【Abstract】 ObjectivesEndometrial cancer is one of the most common cancers in women,with a better prognosis for early-stage patients,with a 5-year survival rate of about 95%.However,if reional lymph node metastasis occurs,the survival rate drops to only 69.4%,LVSI is closely related to lymph node metastasis and is considered a prerequisite for lymqph node metastasis,Clarifying the LVSI status before surergy can provide a basis for determining whether lymph node dissection is necessary for stage Ⅰ endometrial cancer patients.Different studies hold different views on whether LVSI will have an adverse effect on the prognosis of stage Ⅰ endometrial cancer.This study aims to analyze the relevant risk factors of LVSI in stage Ⅰ endometrial cancer,establish a nomogram predictive model for the occurrence of LVSI in stage 1 endometrial cancer,and evaluate the impact of LVSI on the prognosis of stage Ⅰ endometrial cancer.MethodsThis study retrospectively collected clinical and pathological data of 749 patients with stage Ⅰ endometrial cancer who underwent comprehensive staging surgery at our hospital from January 2014 to December 2019.The included 749 patients were randomly assigned to a training group and a validation group in a ratio of 7:3 using SPSS 27.0 software.Based on postoperative pathological results,the patients were divided into LVSI-negative group and LVSI-positive group.Inter-group correlation analysis of clinical and pathological data of patients in the training group was performed,and binary logistic regression analysis was used to determine the independent risk factors for LVSI occurrence in stage Ⅰ endometrial cancer.A nomogram predictive model for LVSI risk in stage Ⅰ endometrial cancer was constructed in the training group using R language software(R 4.2.3).The model was evaluated and validated in both the training and validation groups using ROC curve,calibration curve,and DCA curve.The endpoint events of this study are set as recurrence or death,with the evaluation indicators being disease-free survival(DFS)or overall survival(OS).Single-factor and multifactor Cox proportional hazards regression analysis were used to explore independent prognostic factors for stage Ⅰ endometrial cancer patients.The propensity score matching algorithm(PSM)was employed to balance the clinical baseline characteristics of patients between the LVSI-negative group and LVSI-positive group.Kaplan-Meier survival analysis was conducted to investigate the impact of LVSI on the prognosis of stage Ⅰ endometrial cancer.Results1.Establishment and Validation of the Risk Prediction Model for LVSI in Stage 1 Endometrual CancerThis study included a total of 749 patients with stage Ⅰ endomtrial cancer who underwent comprehensive staging surgery,among which 608 patients were negative for LVSI and 141 were positive for LVSL All patients were randomly assigned to training group and validation group in a 7:3 ratio.The training group consisted of 528 cases,while the validation group had 221 cases.Inter-group correlation analysis and binary logistic regression analysis in the training group revealed that hypertension,CA-125 levels,maximunm tumor diameter,histological type,depth of myometrial invasion,and histologic grade were significantly associated with the occurrence of LVSI in stage Ⅰ endometrial cancer(P<0.05).CA-125>35 U/ml(OR=2.420,95%CI:1.369-4.276,P=0.002),maximum tumor diameter>2cm(OR=2.032,95%CI:1.2143.401,P=0.007),non-endometrioid histological type(OR=4.792,95%CI:2.249-10.209,P<0.0014),histologic grade G3(OR=3.449,95%CI:1.891-6.291,P<0.001),myometrial invasion depth≥1/2(OR=2.408,95%CI:1.354-4.284,P=0.003)were identified as independent risk factors for LVSI occurrence in stage Ⅰ endometrial cancer.A risk prediction model for LVSI in stage Ⅰ endometrial cancer was constructed using CA-125 levels,maximum tumor diameter,histological type,histologic grade,and depth of myometrial invasion as five predictive factors The model evaluated and validated in both training and validation groups using ROC curve,calibration curve,and DCA curve,demonstrating good discrimination,calibration,and clinical utility.2.The impact of LVSI on the prognosis and independent risk factors for stage 1 endometrial cancerThe follow-up time of this study ranged from 22 to 117 months.During the follow-up period,19 patients(2.5%)died due to the disease,and 34 patients(4.5%)experienced disease recurrence.The 5-year overall survival rate was 95.1%,and the 5-year disease-free survival rate was 93.2%.In all stage Ⅰ endometrial cancer patients included in the study,the results of univariate and multivariate Cox proportional hazards regression analysis showed that histological grade G3(HR=3.441,95%CI:1.296-9.134,P=0.013)and positive LVSI(HR=3.503,95%Cl:1.2429.883,P=0.018)were independent risk factors for DFS in stage Ⅰ endometrial cancer.Histological grade G3(HR=6.404,95%Cl:1.798-22.817,P=0.004),myometrial invasion depth≥1/2(HR=4.036,95%CI:1.101-14.788,P=0.035),and positive LVSI(HR=5.151,95%Cl:1.346-19.710,P=0.017)were identified as independent risk factors OS in stage Ⅰ endometrial cancer patients.Propensity score matching(PSM)was used to balance the intergroup characteristics between patients with negative and positive LVSI.In the unmatched cohort,there were statistically significant differences between the LVSInegative group and the LVSI-positive group in terms of DFS(P<0.001)and OS(P=0.002).In the matched cohort,the differences between the two groups in terms of DFS(P=0.037)and OS(P=0.043)remained statistically significant.Conclusions1.CA-125>35U/ml,tumor maximum diameter>2cm,pathological type is nonendometrioid carcinoma,histologial grade is G3,and deep myometrial invasion ≥ 1/2 is an independent risk factor for lymphovascular space invasion in stge Ⅰ endometrial cancer.2.The risk prediction model for LVSI in stage Ⅰ endometrial cancer developed in this study has good discriminative ablity,calibration,and clinical utility.It can assist clinicians in preliminarily predicting the risk of LVSI in patients with stage Ⅰ endometrial cancer before surgery,aiding in the development of more precise surgical approaches and individualized treatment plans.3.LVSI positivity is an independent poor prognostic factor for stage Ⅰ EC,which will have a negative impact on both DFS and OS.更多还原