Imaging Comparison for the Preoperative Planning and Diagnosis of DIE: a Multicenter Retrospective Study.

Overview

Several centers in Belgium use both transvaginal ultrasound (TVS) and magnetic resonance imaging (MRI) for the preoperative diagnosis of deep infiltrating endometriosis (DIE), while other centers rely almost exclusively on TVS. From the perspective of both the patient and the endometriosis care team, it is not primarily important that every individual lesion is mapped perfectly preoperatively, but rather that all lesions impacting clinical management and surgical planning are accurately detected. This is particularly crucial when a multidisciplinary approach is required, involving a urologist for bladder lesions and/or an abdominal surgeon for invasive rectosigmoid lesions. Moreover, providing the patient with thorough preoperative counseling is essential, and this is, of course, determined by the preoperative findings and the type of planned surgical procedure. In this study, we first assess the diagnostic performance of TVS in the preoperative diagnosis of DIE. As a secondary objective, we evaluate the added value of MRI compared to TVS for preoperative surgical planning in patients who also underwent an MRI examination.

Endometriosis is a common gynecological disease affecting 5 to 10% of women of reproductive age (1,2). Seventy percent of patients are symptomatic (3). This chronic and inflammatory condition is defined as the presence of endometrial cells outside the uterine cavity (1,2). The main symptoms include dysmenorrhea, chronic pelvic pain, dyspareunia, and infertility (1,3). In the literature, three different subtypes of endometriosis are described based on location: superficial peritoneal endometriosis, ovarian endometriosis, and deep infiltrating endometriosis (DIE) (4). DIE can occur in various structures, such as the uterosacral ligaments (USL), retrocervical region, rectovaginal septum (RVS), vagina, round ligament, vesico-uterine septum, bladder, intestines (mainly sigmoid colon, rectum, appendix, and cecum), and ureters (5). Until a few years ago, the average time between the onset of symptoms and diagnosis was between 4 and 11 years (2). Two major reasons for this delay were the minimization of complaints by physicians and the difficulty of diagnosis. Diagnostic laparoscopy, with or without histological confirmation, was then considered the gold standard for diagnosing the disease (6). However, due to the invasive nature of the procedure, it was rarely performed or only after years of persistent symptoms (6). Non-invasive procedures such as clinical examination and imaging were often introduced only later, after diagnostic laparoscopy and before therapeutic surgery (7). The two most commonly used imaging techniques are transvaginal ultrasound (TVS) and magnetic resonance imaging (MRI) (8). When these techniques are used early in the diagnostic process, the time to diagnosis of endometriosis can be significantly shortened (9). As recently emphasized by the consensus statement from the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) (8): "Reducing diagnostic delay and adequately managing endometriosis requires a shift from a surgical or lesion-centered diagnosis to a broader diagnostic approach, taking into account symptoms, signs, and non-invasive findings from clinical examination and imaging." Diagnostic laparoscopy is now only recommended by the European Society for Human Reproduction and Embryology (ESHRE) when imaging results are negative and/or medical treatment fails (10). Recent meta-analyses show that the specificity and sensitivity of TVS are generally comparable to those of MRI for the preoperative diagnosis of DIE (11-14). In detecting rectosigmoid lesions, a systematic review and meta-analysis by Gerges et al. demonstrated that rectal endoscopic ultrasound has the highest sensitivity (92%) for lesion detection (15). Nevertheless, TVS, with a sensitivity of 89%, is recommended as the first-line modality due to its ease of use and broad availability (15). Furthermore, rectal ultrasound cannot provide complete staging of all potential endometriosis locations, necessitating additional TVS (15). Another systematic review and meta-analysis by Moura et al., published in 2019, which exclusively compared TVS and MRI in rectosigmoid lesions, concluded that both techniques have comparable sensitivity (90%) and specificity (96%), without significant differences in diagnostic performance. Moreover, combining TVS and MRI further improves diagnostic accuracy, achieving an almost perfect post-test probability of 99.6% (12). When comparing the diagnostic accuracy of MRI for DIE to the performance of TVS, clear differences emerge depending on the location of the lesions (16,17). MRI generally has a higher sensitivity than TVS in diagnosing endometriosis of the USL and the RVS. At the USL, MRI correctly detects 77% of cases compared to 49% for TVS. For the RVS, this difference is even more pronounced, with a sensitivity of 85% for MRI versus 53% for TVS. Thus, MRI appears better suited for identifying DIE at these locations. On the other hand, TVS shows higher specificity at certain locations, such as the USL and the vagina. For the USL, TVS specificity is 93% compared to 80% for MRI. For the vagina, the difference is even greater, with a specificity of 96% for TVS and 82% for MRI. This indicates that TVS is more effective at avoiding false-positive diagnoses at these sites. For bladder lesions, both techniques perform similarly: MRI has a slightly higher sensitivity (64% versus 62% for TVS), while both offer excellent specificity. MRI achieves a specificity of 98%, while TVS reaches 100%. In summary, MRI is particularly useful for accurately identifying DIE at locations such as the USL and RVS, while TVS is superior in excluding false positives, especially at the USL and the vagina. For bladder lesions, both modalities provide similarly reliable results (16,18). Three guidelines aim to improve and standardize the quality of diagnosis and reporting of endometriosis. The first is the IDEA consensus (International Deep Endometriosis Analysis), published in 2016 by Guerriero et al. on behalf of the IDEA group. It is an internationally accepted guideline describing a stepwise methodology for using TVS to assess all potential locations of endometriosis and to describe lesions in a standardized manner (17). The second is the dPEI classification (Deep Pelvic Endometriosis Index), developed by Thomassin-Naggara et al. and published in 2020 (19). This is a validated MRI classification system specifically designed for the evaluation and preoperative planning of deep pelvic endometriosis (19). Although the system is clinically highly valuable, especially because it includes lateral and extra-pelvic endometriosis locations, it is not yet explicitly recognized as an internationally accepted guideline (19). Finally, the #ENZIAN classification system, published by Keckstein et al. in 2021, offers an internationally accepted system to clearly map all intraoperative endometriosis locations. While this system is more comprehensive than previous classifications, it does not fully account for lateral pelvic wall involvement, which may limit its application for preoperative planning. Both the IDEA consensus and the #ENZIAN classification system are internationally accepted and, together with the dPEI classification, contribute to improving research through standardized methodologies and a uniform language, making results more comparable and interpretable (17,19,20). The internationally accepted consensus statement by ISUOG, published by Condous et al. in 2024, provides guidelines for the use of non-invasive imaging techniques for the diagnosis and classification of endometriosis (8). It emphasizes that TVS is generally reliable for the preoperative diagnosis of DIE but that it is less accurate in detecting higher, more proximal lesions, such as sigmoid DIE, due to TVS's limited field of view. MRI is valuable in predicting the presence of DIE in the RVS and offers a more reliable diagnosis of bladder endometriosis (8). However, the longer waiting time, higher cost, and greater environmental impact of MRI compared to TVS (21) are important arguments for recommending TVS, performed by an expert operator, as the primary preoperative diagnostic tool for DIE without routine MRI use (8). Several centers in Belgium use both TVS and MRI for the preoperative diagnosis of DIE, while other centers rely almost exclusively on TVS. From the perspective of both the patient and the endometriosis care team, it is less important to map each individual lesion perfectly preoperatively than it is to detect all lesions that will impact clinical management and surgical planning. This is particularly crucial when a multidisciplinary surgical intervention is necessary, involving a urologist for bladder lesions and/or an abdominal surgeon for invasive rectosigmoid lesions. Moreover, providing the patient with thorough preoperative counseling is vital, and this counseling is of course determined by the preoperative findings and the type of planned surgical procedure. In this study, we first evaluate the diagnostic performance of TVS for the preoperative diagnosis of DIE. As a secondary objective, we assess the added value of MRI compared to TVS for preoperative surgical planning in patients who also underwent an MRI examination.