国际生殖健康/计划生育杂志

国际生殖健康/计划生育杂志 ›› 2026, Vol. 45 ›› Issue (5): 368-372.doi: 10.12280/gjszjk.20260383

• 论著 • 上一篇    下一篇

GnRH拮抗剂方案新鲜胚胎移植妊娠结局的列线图预测模型构建

李欣欣, 黄雨姗, 干恬, 孟悦()   

  1. 510630 广州， 中山大学附属第三医院生殖医学中心（李欣欣， 黄雨姗，干恬），妇科（孟悦）
  • 收稿日期:2026-07-17 出版日期:2026-09-15 发布日期:2026-09-20
  • 通讯作者: 孟悦，E-mail：mengy67@mail.sysu.edu.cn

Construction of A Nomogram Model for Predicting Pregnancy Outcomes in Fresh Embryo Transfer Cycles Using GnRH Antagonist Protocols

LI Xin-xin, HUANG Yu-shan, GAN Tian, MENG Yue()   

  1. Center for Reproductive Medicine (LI Xin-xin, HUANG Yu-shan, GAN Tian), Department of Gynecology (MENG Yue), The Third Affiliated Hospital of Sun Yat-Sen University, Guangzhou 510630, China
  • Received:2026-07-17 Published:2026-09-15 Online:2026-09-20
  • Contact: MENG Yue, E-mail: mengy67@mail.sysu.edu.cn

RichHTML

11

PDF (PC)

80

摘要：

目的：探讨促性腺激素释放激素（gonadotropin-releasing hormone，GnRH）拮抗剂方案新鲜胚胎移植临床妊娠率的影响因素，并构建及验证预测模型。方法：回顾性分析2018年1月—2024年12月在中山大学附属第三医院进行拮抗剂方案新鲜胚胎移植的2 405例患者，按7∶3随机分为训练集（1 683例）与验证集（722例）。通过多因素Logistic回归分析筛选临床妊娠的独立影响因素并建立列线图。模型性能采用受试者工作特征曲线下面积（area under curve，AUC）及校准曲线进行评价。结果：多因素Logistic回归分析表明，D3优质胚胎数增加（OR=0.88，95%CI：0.82~0.94）及移植胚胎数增加（以移植1枚胚胎为对照，移植2枚胚胎OR=0.52，95%CI：0.37~0.72）发生未临床妊娠的概率降低；女方年龄增加（OR=1.04，95%CI：1.00~1.09）、扳机日孕酮升高（OR=2.44，95%CI：1.46~4.09）发生未临床妊娠概率增加。基于这4项独立预测因子，建立临床妊娠预测模型并绘制列线图。训练集AUC为0.64（95%CI：0.61~0.67），验证集AUC为0.65（95%CI：0.61~0.69），模型区分度中等。训练集和验证集的校准曲线显示预测概率与实际概率高度一致，提示模型校准度良好。结论：构建的列线图具有较好的性能，可试用于临床妊娠率的个体化预测，为患者提供基于数据的胚胎移植策略建议。

关键词: 生殖技术, 辅助, 胚胎移植, 促性腺素释放激素, 妊娠结局, 列线图, GnRH拮抗剂方案

Abstract:

Objective: To investigate the factors influencing the clinical pregnancy rate in fresh embryo transfer cycles using gonadotropin-releasing hormone (GnRH) antagonist protocols, and to construct and validate a predictive model. Methods: A total of 2 405 fresh embryo transfer cycles using GnRH antagonist protocols were retrospectively analyzed, at the Third Affiliated Hospital of Sun Yat-Sen University between January 2018 and December 2024. The cycles were randomly allocated into a training cohort (n=1 683) and a validation cohort (n=722) at a 7∶3 ratio. The independent factors influencing clinical pregnancy were identified by multivariate logistic regression analysis, which were subsequently used to construct a nomogram. The model performance was then evaluated using the area under the receiver operating characteristic curve and calibration curves. Results: Multivariable logistic regression analysis revealed that a greater number of high-quality D3 embryos (OR=0.88, 95%CI: 0.82-0.94) and the transfer of two embryos (versus one embryo, OR=0.52, 95%CI: 0.37-0.72) were associated with decreased probability of non-clinical pregnancy. Conversely, the advancing female age (OR=1.04, 95%CI: 1.00-1.09) and the elevated serum progesterone on trigger-day (OR=2.44, 95%CI: 1.46-4.09) were associated with increased probability of non-clinical pregnancy. Based on these four independent predictors, a predictive model for clinical pregnancy was developed, and a nomogram was constructed. The model was demonstrated the moderate discriminative performance, with an area under the receiver operating characteristic curve of 0.64 (95%CI: 0.61-0.67) in the training set and 0.65 (95%CI: 0.61-0.69) in the validation set. Calibration curves for both datasets showed a close agreement between predicted and observed probabilities, indicating good model calibration. Conclusions: The developed nomogram demonstrates favorable predictive performance, which may be used for the individualized prediction of clinical pregnancy, thereby providing data-driven recommendations for embryo transfer strategies.

Key words: Reproductive techniques, assisted, Embryo transfer, Gonadotropin-releasing hormone, Pregnancy outcome, Nomograms, GnRH antagonist protocol