Zhongguo aizheng zazhi (Apr 2024)

Magnetic resonance imaging for distinguishing gastric-type endocervical adenocarcinoma from lobular endocervical glandular hyperplasia

  • MA Fenghua, JIANG Anqi, CHEN Yiqing, XU Congjian, KANG Yu

DOI
https://doi.org/10.19401/j.cnki.1007-3639.2024.04.005
Journal volume & issue
Vol. 34, no. 4
pp. 380 – 388

Abstract

Read online

Background and purpose: Gastric-type endocervical adenocarcinoma (G-EAC) is a rare variant of endocervical adenocarcinoma, characterized by atypical clinical manifestations and elusive lesions. Due to these factors, G-EAC is prone to being missed or misdiagnosed, significantly impacting the prognosis. Lobular endocervical glandular hyperplasia (LEGH) and atypical LEGH (aLEGH) are considered to be precancerous lesions of G-EAC. These conditions also present overlapping clinical, pathologic and imaging manifestations, making it challenging to differentiate between them preoperatively. The purpose of this study was to investigate the correlation between magnetic resonance imaging (MRI) findings of cystic-solid lesions in the cervix and their underlying pathology in order to enhance the accuracy of distinguishing between LEGH and G-EAC, ultimately aiding in the early diagnosis and appropriate management of these conditions. Methods: Clinical, imaging and pathological data of 37 LEGH and 53 G-EAC patients who attended the Obstetrics and Gynecology Hospital of Fudan University from July 2016 to August 2023 were collected. Analysis was conducted using Pearson Chi-square χ2, Fisher’s exact tests and so on. Multivariate analyses were performed using logistic regression. Receiver operating characteristic (ROC) curves were used for performance evaluation. Results: In this study, differences in age, symptoms, extent, size, composition, degree of enhancement, cervical stromal ring, endometrium invasion, pelvic lymph nodes enlargement, and hydrohystera were statistically significant between the two groups (P<0.05). In the LEGH and aLEGH groups, lesions were primarily localized to the epithelial layer of the endocervical canal. These lesions were predominantly simple cystic (32/37), and the cystic walls often displayed significant enhancement (31/37). In contrast, the G-EAC group presented with lesions involving the myometrium of the uterine cervix (42/53). These lesions were characterized by a solid component in the majority of cases (52/53), a tendency for the disappearance of the cervical stromal ring (46/53). Logistic regression analysis revealed that among the MRI features, lesion composition (OR = 50.064) and incomplete cervical stromal ring (OR = 40.180) were significant predictors for G-EAC. ROC analysis, incorporating lesion size, composition, enhancement degree, cervical stromal ring, and endometrial involvement, yielded an area under curve (AUC) of 0.970 (95% CI: 0.931-1.008). Conclusion: Combining multiple MRI features of cystic-solid lesions in the cervix aids in distinguishing between LEGH and G-EAC.

Keywords