The role of three-dimensional MRI in the differentiation between angular pregnancy and interstitial pregnancy

Ectopic pregnancy (EP) is the most common cause of pregnancy-related mortality in the first trimester [1]. Interstitial pregnancy is a subtype of EP, with the implantation of the embryo in the intramural or interstitial portion of the fallopian tube. Although interstitial pregnancy only accounts for 1–3% of all EP, its mortality rate is as high as 2.5%, which is 7 times greater than the overall mortality rate in EP [2, 3]. The angular pregnancy is an eccentric intrauterine pregnancy in which embryonic tissue implants in the endometrium along the lateral edge of the uterus, medial to the utero-tubal junction. The embryo may develop or abort in the uterine cavity [4], which means that angular pregnancy is an obscure entity and the outcomes are variable [5]. Patients may suffer from spontaneous abortion or uterine rupture, while some women may eventually achieve successful live births through careful observational management [6]. Due to the different causes, treatments and outcomes, timely and accurate diagnosis of angular pregnancy and interstitial pregnancy is of great significance in clinical practice.

Diagnostic laparoscopy, as a traditional method to diagnose angular pregnancy or interstitial pregnancy under direct vision, is invasive [1, 5]. Ultrasound is currently the most preferred screening tool to evaluate the site of pregnancy implantation worldwide. However, the differences between interstitial pregnancy and angular pregnancy are very subtle sometimes, and the sensibility of the conventional ultrasound is too low to distinguish them [7]. Further, although the three-dimensional (3D) ultrasound is regarded as the best way to show the entire uterine cavity, it is possible to comprehensively evaluate the relation between the gestational sac (GS) position and the uterine cavity [8, 9]. Also, it is worth noting that the experience of the sonographer plays a vital role in the performance of 3D ultrasound, which may limit the use of 3D ultrasound. In recent years, magnetic resonance imaging (MRI) has shown great advantages in the diagnosis of various forms of EP [10, 11]. The role of MRI in diagnosing interstitial pregnancy has been described in a few case reports, but the MRI image analysis criteria has not been comprehensively addressed [1, 12‐15]. Also, the two-dimensional T2-weighted (T2W) sequence plays a significant role in EP cases in the current clinical practice, but it requires multiple planes [10, 13]. With the fast-imaging speed and multi-directional reconstruction capabilities, the 3D T2W images can be effectively used for the reconstruction and identification of uterine anomalies [16, 17]. However, very few studies have reported the role of 3D MRI in the EP diagnosis, especially for angular and interstitial pregnancy.

In this study, our hypothesis is that 3D T2W MRI can provide additional information to distinguish between angular and interstitial pregnancy effectively. We retrospectively reviewed the 3D images of angular pregnancy and interstitial pregnancy, and further investigated MRI features of the 3D images between these two conditions. The purpose of the study is to explore the role of 3D T2W MRI in distinguishing angular pregnancy from interstitial pregnancy in the early diagnosis and management.

Ultrasound is the initial imaging modality for evaluating patients with suspected interstitial pregnancy or angular pregnancy [19, 20]. Previous studies have provided the sonographic criteria for the general diagnosis of interstitial pregnancy and angular pregnancy [6, 21]. It is reported that the “interstitial line sign” has high sensitivity (80%) and specificity (98%) in the diagnosis of interstitial pregnancy [7]. Furthermore, 3D ultrasound makes it possible to delineate the uterine cavity and the intramural portion of the fallopian tube [8]. Also, 3D ultrasound is able to reveal a more confident location of the GS [22]. Edward et al. demonstrated that the niche mode of 3D ultrasound can clearly show the relation between the GS and the uterine cavity, which effectively informed the planning of the therapeutic proposal [23]. Therefore, ultrasound is essential in the diagnosis of angular pregnancy and interstitial pregnancy. However, sometimes the differences between angular pregnancy and interstitial pregnancy are very subtle. Indeed, MRI is not easily performed to investigate abnormal pregnancy due to the high cost and long queuing time for appointments. However, considering the unparalleled advantages and reliability of MRI in distinguishing between angular and interstitial ectopic pregnancies, it is of great value and significance to utilize MRI in the investigation of theses abnormal pregnancies.

MRI has served as a problem-solving procedure in ectopic pregnancy [24]. With the excellent soft tissue resolution, the large field of view and the multi-planar imaging capability, MRI has proved to be a useful tool in diagnosing ectopic pregnancies [10]. However, once the examination is finished, the conventional 2D MRI can only display limited planes. Sometimes more than three planes are required for accurate diagnosis [25], and it is difficult for technologists to determine the optimal choices of imaging planes. 3D MRI allows to construct on any desired plane after examination and does not require a precise planning of planes by the technologist. Therefore, it can add additional information to help diagnose in some cases [25, 26]. In this study, we demonstrate that 3D MRI is helpful in the differential diagnosis between the interstitial and angular pregnancy. Three significant 3D MRI imaging features were found to differentiate between these two pregnancies, including the junctional zone involvements, the surrounding signal intensity, and the outline of uterus cavity. Moreover, a final model combining these features showed a high diagnostic accuracy for both observers.

Three significant 3D MRI imaging features were included in the logistic regression model. The junctional zone is the innermost myometrium adjacent to the endometrium [27], which is best to be visualized on T2WI for evaluation and appears as a low signal region within the innermost myometrium [28]. The interruption of the junctional zone and the myometrial invasion by placental tissue suggest angular pregnancy [29]. Filhastre et al. reported two cases of interstitial pregnancy with an uninterrupted junctional zone between the gestational sac and the uterine cavity, and claim this signals interstitial pregnancy [13]. Consistent with the former study, we found that the involvement of junctional zone was a crucial imaging features to identify angular pregnancy. The junctional zone is a low signal region on the 3D T2W images, which is distinct from the high intensity signal of the endometrium and the intermediate intensity signal of the myometrium. The interruption of the junctional zone can be clearly displayed after the 3D MR image reconstruction, but sometimes this image feature is inconclusive on the orthogonal planes.

The ultrasound sign of “surrounded by endometrial” is a specific sign for angular pregnancies, with 100% specificity according to a retrospective study [19]. On the T2W image, the endometrium is usually high-intensity while the myometrium is low-intensity. Previous studies have shown that a sac surrounded by the endometrium indicates an angular pregnancy, while a sac surrounded by the myometrium indicates an interstitial pregnancy [29]. Consistent with previous studies, we demonstrate that being surrounded by hyper/iso-intensity on 3D T2WI is a hallmark for the diagnosis of angular pregnancy. However, we misdiagnosed 35 cases using this imaging feature, where all these 35 patients had a bigger size of GS-structure and more advanced gestational age. This was possibly because the decidua capsularis abuts the decidua parietalis, and the space separating them was obliterated at advanced gestational age.

The endometrial cavity is the implantation site of angular pregnancy, which could easily cause the asymmetric cavity [1]. However, the conventional ultrasound is hardly able to provide a whole view of the uterus. MRI, especially 3D imaging, allows a more accurate identification of the outline of the uterus cavity [29]. The specificity of this imaging feature reached as high as 80% in our study.

Although a myometrial mantle measurement less than 5 mm and located lateral to the round ligament have been reported as two important indicators of interstitial pregnancy [1, 30], we excluded these two image features in our diagnostic model. Similar to the results of our study, Tulandi and Al-Jaroudi argue that a thin myometrial mantle is not an important sign to distinguish the interstitial pregnancy from an eccentrically located intrauterine pregnancy [31]. We believed it is because large eccentric intrauterine pregnancy can compress the overlying myometrium and the myometrial mantle measurement can be subjective sometimes. In the angular pregnancy, the lateral uterine enlargement of the gestation displaces the round ligament reflection upward and outward, and the gestational swelling is seen medial to the round ligament [32]. However, the sensitivity of this image feature is not as high as expected in our study. We thought it is due to the bowel movement and the lack of contrast, which makes it difficult to identify the round ligament in some patients.

In addition, the use of 3D MRI could also potentially reduce the examination time. The screening time in our study was around 4 min for the 3D CUBE T2W imaging and around 2 min for each plane of the 2D T2W TSE imaging. We obtained three sequences of 2D T2W imaging for each case, so the total imaging time for the conventional 2D T2W imaging was approximately 50% longer than that of the 3D Cube T2W imaging. This result is similar to the findings of previous reports that the 3D MRI can yield time savings up to 50–60% compared with standard 2D-T2WI acquisition in three orthogonal planes [25, 26].

Although the pelvic MRI was taken in the first trimester, no adverse reactions were found in the live-born babies. Fetal cells rapidly proliferate, divide, and then undergo organogenesis in the first trimester. The implantation, migration, or differentiation of fetal cells may be disrupted due to heating theoretically. But no specific consequences of fetuses’ exposure to non-contrast MRI during the first trimester have been documented [33, 34]. Non-contrast MRI is considered safe during pregnancy by the American College of Radiology and the American College of Obstetricians and Gynecologists [35]. Andre L. Chartier’s study showed that no adverse effects regarding neonatal hearing or fetal growth were found for those who were variably exposed to 3-T MR in utero by MRI at any gestational age [36]. Considering the risk-to-benefit ratio, we suggest that the non-contrast MRI should be performed to help distinguish interstitial pregnancy from angular pregnancy.

It is acknowledged that there are some limitations in this study. First, this is a retrospective study. The inclusion criteria may result in a selection bias. Further studies may explore a cohort study to reduce the bias. Second, the sample size is relatively small and all samples are from one single institution which may hide or amplify some clinical features. A larger number of samples are expected to be collected from multi-hospital and different models should be explored using consolidated data in future.

In conclusion, our results demonstrate that 3D T2W images of MRI play an important role in distinguish angular pregnancy from interstitial pregnancy. The 3D T2W image features identified on the reconstruction images, including the junctional zone involvement, asymmetric outline of uterus cavity and being surrounded by hyper/iso-intense, can strongly suggest the angular pregnancy. The combination of the above three significant imaging features can improve the diagnostic performance. Ultimately, 3D MRI enables more accurate characterization of the pregnancies for diagnosis, which can effectively avoid unnecessary interventions, prevent hemorrhage, and help to improve the diagnosis timing and patient management.