摘要
目的分析病理性乳头溢液患者的临床特征及其癌变的影响因素,并构建癌变风险预测模型。方法回顾性分析114例病理性乳头溢液患者的临床资料,总结其临床特征。根据病理检查结果将患者分为癌变组(n=14)和非癌变组(n=100),分析影响病理性乳头溢液患者发生癌变的独立危险因素。构建病理性乳头溢液患者发生癌变的风险预测模型,并采用受试者工作特征曲线评估该模型的预测效能。结果(1)114例病理性乳头溢液患者中,有14例发生癌变,癌变率为12.28%。患者年龄主要集中在40~50岁,有91.23%的患者伴有乳腺肿块,单侧溢液、浆液性溢液更为常见。(2)体质指数>24 kg/m^(2)、血性溢液、溢液时间长、合并糖尿病及血小板计数高是病理性乳头溢液患者发生癌变的独立危险因素(均P<0.05)。(3)构建的风险预测模型为:Y=46.511+1.229×体质指数+1.582×血性溢液-0.335×溢液时间+1.386×合并糖尿病-0.158×血小板计数。该模型预测病理性乳头溢液患者发生癌变的曲线下面积为0.891,大于单一危险因素的曲线下面积(均P<0.05),且灵敏度、特异度、准确率均高于80%。结论病理性乳头溢液患者高发年龄为40~50岁,常伴有乳腺肿块,以单侧溢液或浆液性溢液为主。而体质指数>24 kg/m^(2)、血性溢液、溢液时间长、合并糖尿病及高血小板计数的患者癌变风险更高,此类患者应及时就诊。根据上述危险因素构建的癌变风险预测模型对临床上指导病理性乳头溢液患者降低癌变风险具有一定作用。
Objective To analyze the clinical features of pathological nipple discharge patients and the influencing factors of cancerization,and to establish a risk prediction model for cancerization.Methods The clinical data of 114 patients with pathological nipple discharge were analyzed retrospectively,and the clinical features were summarized.According to the pathological results,the patients were divided into cancerization group(n=14)and non-cancerization group(n=100),and the independent risk factors affecting cancerization in patients with pathological nipple discharge were analyzed.A risk prediction model for cancerization in patients with pathological nipple discharge was established,and the predictive efficiency of the model was assessed by using receiver operating characteristic curve.Results(1)Among 114 patients with pathological nipple discharge,14 cases suffered from cancerization,and the rate of cancerization was 12.28%.The majority of patients aged 40~50 years,the cases accompanied by breast mass accounted for 91.23%,and unilateral discharge and serous discharge were more common.(2)Body mass index(BMI)>24 kg/m^(2),bloody discharge,longer duration of discharge,concomitant diabetes,and high PLT count were the independent risk factors for cancerization in patients with pathological nipple discharge(all P<0.05).(3)A risk prediction model had been established:Y=46.511+1.229×BMI+1.582×bloody discharge-0.335×discharge duration+1.386×concomitant diabetes-0.158×platelet count.For predicting cancerization in patients with pathological nipple discharge,the model obtained an area under the curve(0.891)higher than that of any of the risk factors(all P<0.05),and possessed sensitivity,specificity and accuracy higher than 80%.Conclusion Pathological nipple discharge is prevalent among the patients aged 40-50 years,commonly accompanied by beast masses,and dominated by unilateral or serous discharge.Patients with BMI>24 kg/m^(2),bloody discharge,long duration of discharge,concomitant diabetes,or high PLT count are at high
作者
卢小华
吴永晓
张涛
邱梅婷
黄临凌
谢凯圣
杨凤英
廖柳芳
LU Xiao-hua;WU Yong-xiao;ZHANG Tao;QIU Mei-ting;HUANG Lin-ling;XIE Kai-sheng;YANG Feng-ying;LIAO Liu-fang(Department of Breast Surgery,the Second Affiliated Hospital of Guangxi University of Science and Technology,Liuzhou 545000,China)
出处
《广西医学》
CAS
2021年第15期1816-1821,共6页
Guangxi Medical Journal
基金
广西壮族自治区卫生健康委员会自筹经费科研项目(Z20190607,Z20180136,Z2016718)。
关键词
乳头溢液
临床特征
癌变
危险因素
预测模型
Nipple discharge
Clinical characteristics
Cancerization
Risk factor
Prediction model
