闭经患者孕激素撤退性出血100例临床分析

Clinical analysis of 100 menopause patients with progesterone withdrawal bleeding

目的探讨停经患者子宫内膜厚度(EM)和雌激素水平(E_2)预测孕激素撤退性出血的价值。方法收集100例停经和继发闭经患者,利用阴道超声测量EM和化学发光法测量E2水平,然后应用受试者工作特征曲线(ROC)分析肌注黄体酮后撤退性出血的资料。结果 (1)EM和E_2预测孕激素撤退性出血的R(OC曲线下面积AUC1及AUC2分别为0.73及0.79,均大于机会参考线下面积(P<0.05),AUC2大于AUC1,但无统计学差异(P>0.05)。(2)根据ROC曲线计算,EM的最优截断值为0.5 cm,预测孕激素撤退性出血的灵敏度为75.5%,特异度为60.0%,阳性预测值为67.0%,阴性预测值为46.6%;E_2最优截断值结合临床选择为146.4 pmol/L,预测孕激素出现撤退性出血的灵敏度为74.7%,特异度为57.0%,阳性预测值为70.0%,阴性预测值为62.0%。(3)EM的截断值为0.5 cm及以下者孕激素撤退出血阳性者18例(40.0%),E_2截断值为146.4pmol/L以下者孕激素撤退出血阳性者21例(42.0%);EM联合E_2预测孕激素撤退性出血的灵敏度为92.0%,特异度为70.8%,阳性预测值为76.7%,阴性预测值为89.5%。结论 EM和E_2对预测停经患者孕激素能否引起撤退性出血具有临床价值,EM和E_2联合可提高预测的灵敏度及特异度。但在临床工作中,EM和E_2均无确定的阈值,孕激素撤退实验是检验患者是否有内源性雌激素分泌的直接方法,不能仅凭雌激素水平或内膜厚度判断。

Objective： To evaluate the value of endometrial thickness （EM） and estrodial （E2） levels in predicting progesterone withdrawal bleeding in amenorrhea patients. Methods： Endometrial thickness and serum E2 levels were measured by vaginal B ultrasound and chemiluminescence respectively before progesterone withdrawal test in 100 secondary amenorrhea patients from the reproductive endocrinology outpatient clinic of Peking Union Medical College Hospital. The predicting value of endometrium thickness and E2 levels for withdrawal bleeding were analyzed by receiver operator characteristic curve （ROC） . Results： （1） The area under the ROC curve for prediction of progesterone withdrawal bleeding with EM （AUC1） and E2 （AUC2） were 0.73 and 0.79 respectively, which were greater than the opportunity to reference line area （P〈0.05） . Although AUC2 was slightly greater than AUC1, butthere was no significant difference （P 〈 0.05 ） （2） According to the ROC curve calculation, the optimal cutoff value of EM is 5 mm, the sensitivity for progesterone withdrawal bleeding 75.5 %, the specificity 60.0 %, the positive predictive value 67.0 % and the negative predictive value 46.6 %. While the optimal cutoff value of E2 is 40pg/dl （146.4pmol/L）, the sensitivity 74.7%, the specificity 57.0%, the positive predictive value 70.0% and the negative predictive value 62.0%. （3） In the patients with endometrium thickness ≤5 mm, withdrawal bleeding occurred in 18 patients, which was accounted for 40.0%. While in patients with E2 levels≤146.4 pmol/L, withdrawal bleeding occurred in 21 patients, which was accounted for 42.0 %. Taking the account of Ez and EM together, the sensitivity for prediction of progesterone withdrawal bleeding is 92.0G and the specificity 70.8%, the positive predictive value 76.7 % and the negative predictive value 89.5 %. Conclusions： EM and E2 in amenorrhea patients is useful in predicting the results of progesterone withdrawal bleeding to some extent. Combined analysis of EM and E2 improves the sensitivity and specificity of predicting bleeding. Since there are no EM and E2 definite thresholds in clinic practice, the progesterone withdrawal experiment is a direct method to check whether there is endogenous estrogen secretion which can not be judged only by estrogen levels or endometrial thickness gauge.