What to choose and why to use – a critical review on the clinical relevance of rASRM, EFI and Enzian classifications of endometriosis

Stimulated by the fact that the commonly used rASRM classification does not adequately describe DE, the Austrian-German-Swiss Scientific Endometriosis Foundation (Stiftung Endometrioseforschung/SEF) created and published the Enzian classification of endometriosis in 2003. The aim was to better describe and stage DE and to add missing information to the r-ASRM classification ( Keckstein et al., 2003 ; Tuttlies et al., 2005 ). The Enzian classification was revised in 2009 to simplify its use. The rapid development of surgery for DE required a detailed description of the disease to enable comparison of the effects and risks of complications between different surgical techniques and of the accuracy between diagnostic methods. In contrast to the rASRM and EFI classifications, the Enzian classification describes DE involving the vagina, uterosacral ligaments, bladder, ureter, bowel, uterus, and other extragenital localisations. It also takes into account the size of the DE lesions. When using the Enzian classification, the pelvis is divided into three compartments. Compartment A includes the rectovaginal space, vagina, and torus uterinus (cranio-caudal axis), compartment B includes the utero-sacral ligaments, cardinal ligaments, parametrial space and pelvic sidewall (medio-lateral axis), and compartment C includes the rectum and sigmoid colon up to 16 cm from the anal verge (cranio-caudal axis but posterior to compartment A). The grade, i.e. the severity of endometriosis excluding minor peritoneal lesions with less than 5 mm infiltration depth, is defined for each of these three compartment as follows: grade 1 means infiltration 3 cm. Uterine involvement and other extragenital locations of DE (compartment F) are described as adenomyosis (FA), bladder DE (FB), extrinsic and/or intrinsic ureteric involvement with signs of ureteric obstruction (FU), bowel DE cranial to the rectosigmoid junction (FI) (>16 cm from the anal verge; upper sigmoid, transverse colon, caecum, appendix, small bowel), and other locations, e.g. the abdominal wall, diaphragm or nerve/sacral root involvement (FO). Di Paola et al. ( 2015 ) and Burla et al. ( 2019 ) compared MRI findings with surgical findings using the Enzian classification and found good agreement between surgical and MRI Enzian classifications. Thomassin-Naggara et al. ( 2020 ) showed that the Enzian classification based on MRI findings is reproducible and correlates with surgical findings. Their retrospective observational study included 150 patients with DE that underwent MRI and subsequent surgery. The MRI based and surgical Enzian classifications were concordant for DE lesions in the A compartment in 78.7% (118/150), for B lesions in 34.7% (52/150) and for C lesions (colorectal DE) in 82.7% (124/150). Agreement between the radiologists assessing the MR images (inter-observer agreement) was good for DE in the A and C compartments but poor for lesions in the B compartment. Operating times and hospital stays were longer in patients with A2 than A0 lesions, B2 than B0 lesions, C3 than C2 lesions and C2 than C0 lesions according to MRI. This illustrates that there was an association between the Enzian disease grades assessed by MRI and surgical complexity. Patients with vaginal or rectosigmoid involvement (compartment A and C) according to MRI were six and three times more likely to experience grade III Clavien- Dindo complications (i.e., complications requiring surgical, endoscopic, or radiological intervention) than patients without vaginal or rectal DE. Hudelist et al. ( 2019 ) evaluated the lesion location and size according to the Enzian classification by using preoperative TVS. They compared the ultrasound results with surgical findings in 195 women with DE and found good agreement between ultrasound and surgical findings especially for Enzian compartments A, C and FB. Concordance was highest for Enzian compartment C (rectosigmoid), in which 86% of all TVS C3 lesions were confirmed at surgery. Results were similar for Enzian compartment A (vagina, rectovaginal septum). In agreement with results of MRI studies, concordance between TVS findings and surgical findings was poorer for B lesions (uterosacral ligaments, parametria) with 71% of TVS B2 lesions being confirmed at surgery. In most cases of discordant findings, TVS underestimated lesion size by 1 severity grade compared with the intraoperative findings. TVS detected DE in compartments A, B, C, and FB with a sensitivity of 84%, 91%, 92%, and 88%, respectively, and specificity 85%, 73%, 95%, and 99%. Two studies have examined the correlation between the Enzian classification of DE (location, grade) and the severity of preoperative pain symptoms. Both found a correlation ( Haas et al., 2013 ; Montanari et al., 2019 ). Several groups have shown that operating times and risk of surgical complications can be predicted by the Enzian classification. Haas et al. ( 2013 ) demonstrated that intra-operative Enzian classification correlated with the duration of the surgical procedure. To create a prediction model for the risk of surgical complications, Poupon et al. ( 2019 ) developed a nomogram based on three simple criteria; the age of the patient, previous surgery for DE, and the extent of disease described by the surgical Enzian classification. Patients were classified as being at low, intermediate or high risk of surgical complications based on the Enzian classification (low risk defined as A0, A1, B1 and C0, intermediate risk as A2 and/or B2, high risk as A3, B3 or C1). The risk of surgical complications was lowest in the low-risk group and highest in the high-risk group. Imboden et al. ( 2021 ) reported a higher risk of postoperative voiding dysfunction in patients with DE involving the B compartment, especially for B3 lesions. Finally, Nicolaus et al. ( 2020 ) found a statistically significant 3.5-fold increased risk of Clavien-Dindo complications grade II or higher in the presence DE in the Enzian score. Furhermore, an Enzian C3 finding increased the risk of complications greater than Clavien- Dindo grade I 56.3 times (p < 0.001). Based on the above information, several guidelines suggest the use of the Enzian classification for desciption of endometriosis ( Johnson et al., 2017 ; Ulrich et al., 2014 ; Ulrich et al., 2013 ; Vanhie et al., 2016 ). A major criticism of the Enzian classification has been its focus on retroperitoneal, deep infiltrating disease. To overcome this, the #Enzian classification ( Figure 3 ) was created. It is based on consensus between experts after discussions in 2019 and 2020 ( Keckstein et al., 2021 ). The #Enzian classification includes description and classification of peritoneal and ovarian endometriosis and of tubal adhesions, tubal mobility and patency. The novelty of the #Enzian classification is that it takes all the structural manifestations of the disease into account and so allows a complete description of the disease. It can be used both for surgical staging and for staging using imaging, with the exception that superficial peritoneal disease is poorly detectable by current imaging methods and that MRI has limited ability to detect adhesions. The IDEA (International Deep Endometriosis Analysis group) terminology to describe the sonographic appearance of DE ( Guerriero et al., 2016 ) can be incorporated into the #Enzian system. Future studies are needed to test the reproducibility of the #Enzian system when used by surgeons and imaging specialists, the diagnostic accuracy of MRI and TVS when using the #Enzian classification for presurgical staging, and the association of the #Enzian classification with symptoms (including subfertility) and surgical complications. The #Enzian classification now also including peritoneal (“P”), and ovarian endometriosis (“O”) as well as adnexal adhesions (“T”).

A comparison between the rASRM, EFI, Enzian and #Enzian classification systems is provided in Table I . The developments in imaging in the last two decades have not only improved non- invasive diagnosis of endometriosis but also opened a possibility to describe the extent of the disease before surgery (“pre-operative staging”). Although the rASRM classification is the most frequently used system for describing the severity of endometriosis, it does not take into account the retroperitoneal, deep infiltrating phenotypes of endometriosis. The EFI has been extensively validated regarding prediction of pregnancy rates after surgery for endometriosis and appears to be a suitable tool for this purpose. However, like the rASRM, the EFI does not describe DE. Because neither the EFI nor the rASRM classification takes DE into account, they do not reflect the full spectrum of the disease. Therefore, they are not suited to predict the complexity of surgery and risk of surgical complications. Description of the full spectrum of the disease is also needed if one wants to investigate the association between different types of endometriosis and symptoms or the effect of treatment on different types of endometriosis. Comparison of the different classification systems, rASRM, EFI and Enzian including its recently updated and proposed version #Enzian. The advantages and disadvantages of the systems are compared (- not suitable, + little, ++ moderately and +++ well suitable, u.i., under investigation). Until recently, the Enzian classification was predominantly used to describe DE. It is a commonly used method of staging DE in internationally recognised centres of expertise ( Vanhie et al., 2016 ). A great advantage of the Enzian classification is that it can be used to describe the disease on ultrasound or MRI. However, when using ultrasound or MRI the are a few problems. First, MRI measurements of the size of DE affecting the B compartment (uterosacral ligaments, parametrium) are poorly reproducible, and secondly MRI findings of lesions in the B-compartment show poor concordance with surgical findings. Similarly, measurements of parametrial DE by TVS, which have been investigated in only one study ( Hudelist et al., 2021 ) are less reliable (when compared with surgical findings) than measurements of lesions in the A, C and FB compartments. In favour of the Enzian system, there is increasing evidence that the Enzian classification can predict surgical complexity and complications. The #Enzian classification is a step towards the use of one universal classification system, because it includes description not only of DE but also of peritoneal and ovarian endometriosis, adnexal adhesions, and tubal patency. To what extent the #Enzian classification can be used to predict fertility needs to be investigated. We also need studies examining the association between the #Enzian classification and symptoms and between the #Enzian classification and surgical complexity. The inter- and intra-observer reproducibility of the classification when used by surgeons and imaging experts also needs to be investigated. The ideal classification system for endometriosis should describe the sites and extent of the disease, be related to surgical complexity and to disease- associated symptoms, including subfertility. Most importantly, it should be possible to use the classification system both by imaging specialists for pre-operative classification and by surgeons. The need for such a system is obvious and is provided by the #Enzian classification. The validation of this system regarding the above aspects should be the focus of future research.

Since the very first description of endometriosis in 1860 (Rokitansky, 1860), several attempts have been made to classify and describe the anatomical extent of endometriosis and adhesions caused by it. The classification systems have changed over the years, but none has provided a clinically useful system that describes both peritoneal, ovarian, and deep endometriosis (DE). Due to the inadequacy of the existing systems, leading experts suggested that one should use a combination of the three most popular systems, i.e., the revised American Society for Reproductive Medicine (rASRM) classification, the Enzian classification and the endometriosis fertility index (EFI) ( Johnson et al., 2017 ). This is not practical. A correct morphological-anatomical description of endometriosis is a conditio sine qua non for comparing the effect of different therapies of endometriosis and to describe the natural course of the disease. Therefore, one single terminology to describe the location and extent of endometriosis is needed, and this terminology should be the same for imaging experts and surgeons. Today, ovarian endometriomas and DE can be detected by ultrasound or magnetic resonance imaging (MRI), ( Guerriero et al., 2015a ; Guerriero, et al., 2015b ; Guerriero, et al., 2016 ; Gerges et al., 2021 ). Adhesions can also be suspected on transvaginal ultrasound (TVS) ( Gerges et al., 2017 ; Holland et al., 2010 ). However, superficial peritoneal lesions cannot be reliably diagnosed by any imaging method ( Kiesel and Sourouni, 2019 ). As a result, we now can counsel women appropriately before surgery; we can stage endometriosis, predict the complexity of surgery, the risk of surgical complications, and the likely outcome of treatment based on imaging results. The ideal classification system should be useful both for describing disease extent based on imaging results and based on surgical findings. Endometriosis should be managed by a multidisciplinary team, including radiologists, sonographers, gynaecologists, and experts in gynaecological, colorectal, and urological surgery. Therefore, a common language and classification system is needed.