Objectives To retrospectively evaluate the diagnostic value of MRI for the uterotubal junctional pregnancies during the first trimester. Methods This retrospective study involved 59 patients (January 2016 to July 2021) with a preoperative imaging diagnosis of uterotubal junctional pregnancy. Using operative and pathological reports as the reference standard, we identified 22 patients with upper-lateral intracavitary (angular) pregnancy and 37 patients with interstitial pregnancy. Two senior radiologists, blinded to the patients’ information, reviewed the MRI images and determined each MRI feature based on the original interpretation criteria. Any disagreement was resolved by discussion to achieve a consensus. The sensitivity and specificity of each MRI feature were calculated according to the reference standard. Results The endometrial thickness in the upper-lateral intracavitary pregnancy group was larger than in the interstitial group (p = 0.001). The cutoff value of the endometrial thickness was 11.5 mm with a sensitivity, specificity, and area under the curve that were 77.3%, 64.9%, and 0.743, respectively. Two key features to diagnose upper-lateral intracavitary pregnancy were “medial free edge” and “medial free edge plus above-cutoff endometrial thickness.” The sensitivity and specificity of the medial free edge were 100% and 94.9%, respectively. The sensitivity and specificity of the medial free edge plus above-cutoff endometrial thickness were 77.3% and 100%, respectively. The key feature to diagnose interstitial pregnancy was an “intact lateral junctional zone,” of which the sensitivity and specificity were 94.6% and 100%, respectively. Conclusions MRI can be used to differentiate the upper-lateral intracavitary pregnancy and interstitial pregnancy during the first trimester. Key Points • We demonstrated MRI diagnostic criteria for the interstitial pregnancy and upper-lateral intracavitary pregnancy. • MRI might be used to identify the complex interstitial pregnancies, those with a gestational sac protruding into the uterine cavity.

Background The pregnancy outcomes in women with surgery-categorized upper-lateral intracavitary pregnancy (ULIP), previously named angular pregnancy, demonstrate higher heterogeneity than in women with ultrasonography-categorized ULIP. We aimed to use preoperative MRI and correlated clinical characteristics to explore whether the surgery-categorized ULIP comprises obstetric conditions undefined by the current ultrasonography-based diagnostic criteria. Methods This retrospective study involved 28 women with surgically and pathologically confirmed ULIP from January 2016 to July 2022. Two board-certified radiologists, blinded to the patients’ information, independently reviewed the MRI images, and determined each MRI feature, including endometrial thickness (EMT) and peri-gestational sac (GS) endometrial interruption. Disagreements were resolved by discussion to achieve a consensus. Based on the cutoff value of EMT (11.5 mm), the patients were divided into above-cutoff EMT (n = 22) and below-cutoff EMT (n = 6) groups. Results Two subtypes of surgery-categorized ULIP were identified. Type-I ULIP (n = 22; EMT ≥ 11.5 mm), when compared to the type-II ULIP (n = 6; EMT < 11.5 mm), demonstrated lower incidence of peri-GS endometrial interruption (2/22 [9.1%] vs 6/6 [100%]; P = 0.001), higher logarithmic ß-human chorionic gonadotropin (ß-hCG) concentration (4.7 ± 0.4 mIU/ml vs 4.2 ± 0.6 mIU/ml; P = 0.026), lower rate of repeated dilatation and curettage (1/22 [4.6%] vs 4/6 [66.7%]; P = 0.003), less intraoperative blood loss (10.1 ± 6.3 ml vs 28.3 ± 18.3 ml; P = 0.001), and shorter hospital stay (2.8 ± 1.7 days vs 7.5 ± 3.8 days; P = 0.001). The peri-GS endometrial interruption negatively correlated with EMT (Odds ratio [OR] = 0.55; P = 0.001) and logarithmic ß-hCG concentration (OR = 0.08; P = 0.045). The below-cutoff EMT negatively correlated with ß-hCG concentration (OR = 0.06; P = 0.021). Conclusions Surgery-categorized ULIP comprised two obstetric conditions among which the type-II ULIP, possessing unique imaging features undocumented in the literature, requires further attention during clinical practice.