Nomogram based on clinical and DCE-MRI characteristics for predicting the depth of myometrial invasion and grade of endometrioid endometrial carcinoma
10.3760/cma.j.cn112141-20241211-00658
- VernacularTitle:基于临床特征和DCE-MRI定量特征构建的列线图及其预测子宫内膜样癌肌层浸润深度和肿瘤级别的临床价值
- Author:
Xiaoliang MA
1
;
Songqi CAI
;
Jinwei QIANG
;
Guofu ZHANG
;
Jianjun ZHOU
;
Mengsu ZENG
;
Xiaojun REN
;
Rong JIANG
;
Minhua SHEN
Author Information
1. 复旦大学附属中山医院放射科，上海 200032
- Publication Type:Journal Article
- Keywords:
Carcinoma, endometrioid;
Neoplasm invasiveness;
Magnetic resonance imaging;
Nomograms
- From:
Chinese Journal of Obstetrics and Gynecology
2025;60(3):202-215
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To investigate the feasibility and value of nomogram based on base line clinical and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) characteristics for pretreatment predicting the depth of myometrial invasion and tumor grade of endometrioid endometrial carcinoma (EEC).Methods:Preoperative baseline clinical characteristics and DCE-MRI characteristics of 194 EEC patients were prospectively collected at Obstetrics and Gynecology Hospital, Fudan University from October 2020 to January 2022 and used as a training set. Univariate analysis was conducted to compare baseline clinical characteristics and DCE-MRI quantitative parameters [including tumor volume, and mean, median, and standard deviation of volume transfer constant (K trans), rate constant (K ep), extravascular extracellular volume fraction (V e), and initial area under the enhancement curve (iAUC)] between patients with deep myometrial invasion (DMI) and those with superficial myometrial invasion (SMI), as well as between high-grade and low-grade EEC. Multivariate logistics regression analysis was used to identify independent predictors for the construction of nomogram. An independent external testing set comprising 127 EEC patients was retrospectively collected from Zhongshan Hospital, Fudan University and Zhongshan Hospital, Fudan University (Xiamen Branch). The area under the receiver operating characteristic curve (AUC) and decision curve analysis (DCA) were used for evaluating the model′s predictive performance and clinical net benefit, respectively. Results:(1) The depth of myometrial invasion: univariate analysis showed that in the training set, the EEC patients with DMI differed significantly from those with SMI in clinical characteristics including higher proportion of postmenopausal state and overweight [body mass index (BMI)≥25 kg/m2], and abnormal levels of serum cancer antigen (CA) 125, CA 199, and human epididymis protein 4 (HE4), and in DCE-MRI quantitative parameters including tumor volume, and median, mean, and standard deviation of K trans, median of V e, as well as median, mean, and standard deviation of iAUC (all P<0.05). Multivariate analysis showed that the patient′s menstrual status, BMI, CA 199, tumor volume, and mean of iAUC were independent predictors of the depth of myometrial invasion, and constructed the nomogram (recorded as Nomogram_1), achieving an AUC of 0.861 (95% CI: 0.803-0.919) in the training set. In the independent external testing set, the AUC was 0.876 (95% CI: 0.815-0.938), with corresponding sensitivity of 82.0%, specificity of 80.7%, accuracy of 81.1%, positive predictive value (PPV) of 65.3%, and negative predictive value (NPV) of 91.0% for predicting DMI. (2) The EEC grade: univariate analysis showed that in the training set, high-grade EEC patients differed significantly from low-grade EEC in clinical characteristics including patient′s age, the proportion of postmenopausal state and overweight, and abnormal levels of serum CA 125, and in DCE-MRI quantitative parameters including tumor volume, median, mean, and standard deviation of K trans, median and mean of V e, as well as median, mean, and standard deviation of iAUC (all P<0.05). Multivariate analysis showed that the patient′s menstrual status, BMI, tumor volume, and median of V e emerged as independent predictors of EEC grade, and constructed the nomogram (recorded as Nomogram_2), achieving an AUC of 0.845 (95% CI: 0.786-0.893) in the training set. While in the external testing set, the AUC was 0.819 (95% CI: 0.744-0.894), with corresponding sensitivity of 72.4%, specificity of 72.4%, accuracy of 72.4%, PPV of 43.8%, and NPV of 89.9% for predicting high-grade EEC. (3) The DCA curves demonstrated that both Nomogram_1 and Nomogram_2 yielded obvious positive clinical net benefits across a wide range of threshold probabilities. Conclusion:The nomogram based on pretreatment clinical and DCE-MRI characteristics has the potential to noninvasive predict the depth of myometrial invasion and grade of EEC, providing valuable reference information for clinical management decision-making.
