To the best of our knowledge, this study represents the first large-scale prospective observational cohort study to explore the relationship between MRI-detected features of adenomyosis and clinical outcomes of ART, indicated by live birth rates. We found that focal adenomyosis and JZ abnormalities predict live birth rates in adenomyosis.
Most existing studies suggest that women with adenomyosis experience higher miscarriage rates and lower ongoing pregnancy and birth rates[6,20]. In a meta-analysis conducted in 2017, Younes et al. reviewed a total of 18 studies, including nine clinical studies, four prospective cohort studies, and five retrospective cohort studies. Among these studies, six reported that the IVF miscarriage rate was higher in women with adenomyosis, while three studies reported no significant impacts. This discrepancy may stem from the lack of classification or staging of adenomyosis in these studies, as well as differences in the characteristics of the study cohort and the unknown severity of the condition, making it challenging to draw objective and accurate conclusions [6]. Nonetheless, in 2018, Stanekova et al. reported that, after excluding embryo factors, the miscarriage rate in women with adenomyosis was significantly higher than that in those without [20].
Thus, researchers have long been focused on identifying factors and indicators that influence pregnancy outcomes. Accordingly, some studies suggest that JZ thickening in women with adenomyosis negatively impacts pregnancy outcomes. Three distinct layers can be distinguished in the T2-weighted images of women of reproductive age: the endometrium with a high intensity, the outer myometrium with an intermediate intensity, and a low-intensity band in between, referred to as the JZ. The JZ, a unique myometrial band-like structure, was first described by Hricak et al. in 1983 on MRI [21]. A comparison of histological measurements of excised uterine tissue with MRI data revealed that the hyperintense layer on MRI corresponded to the endometrial thickness. Conversely, the low-intensity band only contained the innermost myometrium, characterized by increased cellular density, corresponding to the thickness of the JZ. Subsequently, a more recent study further confirmed the match between the MRI-identified JZ and the histological inner myometrium [22]. Additionally, a retrospective study using the Morphological Uterus Sonographic Assessment (MUSA) criteria to diagnose the outcome of donor assisted pregnancy in patients with adenomyosis showed that uterine volume and diffuse uterine muscle thickening are not related to assisted reproductive outcomes. Instead, the increase in focal lesion diameter is significantly correlated with a decrease in live birth rate, with a decrease coefficient of 0.91; for every 1 cm increase in lesion diameter, the birth rate decreases by 11 times [23]. In contrast to the above results, a prospective observational study on the impact of adenomyosis on in vitro fertilization outcomes in women undergoing donor oocyte transfer in 2024 analyzed the effects of focal adenomyosis, diffuse adenomyosis, or adenomyoma on assisted reproductive outcomes. The results showed that patients with JZ abnormalities had a relative risk of miscarriage that increased by more than three times (relative risk [RR], 3.28; 95% confidence interval [CI], 1.38-7.78). Moreover, the risk of miscarriage doubled in patients with diffuse JZ thickening and severe adenomyosis, while no significant difference was observed in patients with focal adenomyosis [24].
Furthermore, in 2010, Piver et al. explored the relationship between MRI-detected JZ thickness and IVF outcomes, and found a negative correlation between JZ thickness and IVF implantation rates. On this basis, they suggested that MRI-detected JZ thickness could serve as the optimal negative predictor for implantation failure [25]. Thereafter, in recent years, multiple studies have indicated that JZ thickening is associated with implantation failure and mitigated pregnancy rates. After controlling for factors such as advanced age, uterine anomalies, and thin endometrial thickness, studies showed that when the number of transferred embryos, number of high-quality embryos, and pre-transfer endometrial thickness were consistent, the clinical pregnancy rate was significantly lower in the group with a JZmax ≥ 12 mm compared to those with a JZmax < 12 mm [26, 27]. Conversely, the results of this study did not identify JZ thickness as an independent factor influencing the ongoing pregnancy rate or birth rate in women with adenomyosis undergoing assisted reproduction.
Our multiple regression analysis found that focal adenomyosis is one of the reasons affecting the live birth rate of ART in infertile patients. However, unlike previous literature, our study did not find a relationship between the thickness of the JZ and adverse live birth outcomes. Instead, the results of this study revealed that JZ abnormalities on MRI, such as indistinct or difficult-to-observe areas and finger-like indentations, were independent factors influencing live birth outcomes in women with adenomyosis undergoing assisted reproduction.
Most studies and clinicians rely on three main MRI criteria for diagnosing adenomyosis, namely, JZ thickness, JZ-to-myometrium ratio, and the presence of hyperintense regions in the myometrium. However, in 2017, Lisa Agostinho et al. [18] reported several MRI manifestations of adenomyosis beyond these three primary criteria. Additionally, JZ interruption and irregular morphology on 3D ultrasound are listed as ultrasound diagnostic criteria for adenomyosis. Moreover, in 2023, Harmsen et al. [28] described a comparison of MRI and ultrasound features in women with adenomyosis, noting that high-signal intensity spots in the JZ or myometrium on T2-weighted images indicated ectopic endometrial tissue, cystic dilation of endometrial glands, or hemorrhagic lesions. Meanwhile, on transvaginal ultrasound, this appeared as sub-endometrial high-echo lines and buds, JZ interruptions, or hypoechoic cystic areas in the central or external myometrium. Therefore, other JZ abnormalities should be considered for inclusion in the MRI diagnostic criteria in addition to JZ thickening. Thus, the findings of this study expand the scope of MRI diagnostic standards for adenomyosis and provide valuable supporting evidence.
In 2021, an observational cohort study on the MRI manifestations of women with adenomyosis revealed that 109 participants (44.0%) presented exclusively with external adenomyosis. Specifically, the lesion was located in the outer myometrium, separated from the endometrium by an intact JZ and myometrium. Conversely, 78 participants (31.5%) had internal adenomyosis, marked by JZ abnormalities, while maintaining a normal outer myometrium [13]. Owing to the unique anatomical structure of the endometrium, which lacks a submucosal layer for tissue protection, the basal layer and glands are in direct contact with the myometrium. Therefore, repeated endometrial injuries may compromise the integrity of the JZ. Moreover, stromal trophoblasts invade the decidua during embryo implantation. This process triggers initial JZ remodeling, extracellular matrix alterations, uterine smooth muscle relaxation, and spiral artery remodeling, thereby facilitating deep trophoblast invasion [29]. Thus, deficiencies or incomplete remodeling of the inner myometrium are among the contributors to implantation failure. Therefore, in addition to the inflammatory environment caused by adenomyosis, structural changes in the JZ may impair decidualization, leading to inadequate spiral artery remodeling. This disruption in uterine physiology may compromise embryo implantation and successful pregnancy [30].
Moreover, previous research has highlighted a negative correlation between uterine volume and infertility outcomes. A 2023 retrospective cohort study on patients with adenomyosis diagnosed via TVUS undergoing IVF treatment [10] reported that an increased uterine volume was associated with higher miscarriage rates and lower live birth rates. The study identified a uterine volume at eight weeks of gestation (90–130 cm3 for 6–8 weeks of gestation) as a turning point for higher miscarriage rates, as well as a uterine volume at 10 weeks of gestation (130–180 cm3 for 8–10 weeks of gestation, or > 180 cm3 for more than 10 weeks of gestation) as a turning point for lower live birth rates. Additionally, their analysis indicated that patients with a uterus larger than at eight weeks of gestation exhibited significantly higher miscarriage rates, along with lower live birth and cumulative live birth rates across all embryo transfer cycles (fresh and frozen). In contrast, the results of this study did not identify uterine volume as a significant factor influencing pregnancy or live birth outcomes in patients with adenomyosis undergoing assisted reproduction. However, this does not entirely rule out the potential impact of uterine dimensions or JZ thickening on reproductive outcomes. This is because it is possible that patients with larger uterine dimensions in this study may have received more clinical attention, thereby having received appropriate GnRHa pretreatment based on prior studies and clinical experience. Conversely, those with isolated JZ abnormalities identified on MRI, who were perceived as having “mild” conditions, might have been overlooked for GnRHa therapy. Consequently, these findings underscore the need for further randomized controlled trials to explore whether GnRHa pretreatment is beneficial for patients with isolated JZ abnormalities and to determine the optimal pretreatment duration for improving reproductive outcomes.
The strengths of this study include its large-scale and prospective study design. Additionally, our results revealed that focal adenomyosis and abnormal JZs affect the live birth rate of women with adenomyosis. Whereas unlike the results of other studies, no relationship was found between uterine size, JZ thickness, high echogenicity images of the myometrium, and pregnancy outcomes. Notably, previous studies have consistently reported reduced live birth rates in women with adenomyosis. Furthermore, Brosens et al. highlighted that implantation and placental abnormalities were key mechanisms underlying severe pregnancy complications, including late miscarriage, preterm birth, fetal growth restriction, and preeclampsia. These issues are attributed to deep placental damage caused by defective spiral artery remodeling in the JZ [31]. Further randomized controlled trials are warranted to develop targeted interventions for improving pregnancy outcomes in patients with focal adenomyosis and junctional zone abnormalities.
This study had some limitations. First, the participants were selected after a TVS scan and MRI showing signs of adenomyosis, although ultrasound and MRI have good specificity and sensitivity in the diagnosis of adenomyosis. On the other hand, our multiple regression analysis showed no statistical difference in age, the reason for this may be that the included patients were aged 20–38 years, which is relatively young among patients with infertility. Third, we excluded patients with endometrial thickness of <8mm, because we were mainly concerned about the thin endometrium caused by moderate-to-severe intrauterine adhesions, as well as other uterine factors that significantly affect implantation. However, this might have excluded a subset of patients where the adenomyosis affects endometrial proliferation. Further research should include propensity matching to analyze differences in pregnancy outcomes between patients with adenomyosis and those without adenomyosis.