Correlation between findings of nuclear magnetic resonance and 3D anorectal ultrasound in patients with suspected deep endometriosis

Endometriosis Nuclear magnetic resonance 3D anorectal ultrasound Intestinal endometriosis a b s t r a c t

Endometriosis is defined as endometrial glands and stroma that occur outside the uterine cavity. Although not malignant, ectopic endometrial tissue and the result- ing inflammation can cause dysmenorrhea, dyspareunia, chronic pain, and infertility. The diagnostic imaging tests most used are nuclear magnetic resonance imaging (NMR) and ultrasonography (USG).

Correlate the findings of three-dimensional anorectal ultrasound with the NMR findings of the pelvis with intestinal preparation in women with deep endometriosis, through a retrospective, observational, cross-sectional study, evaluating 63 female patients with suspected deep endometriosis with probable involvement intestinal. The evaluation period was from March 2016 to April 2018. Statistical analysis was performed using the kappa agreement to assess the degree of agreement between 3D NMR and USG in relation to the degree of infiltration in the rectal muscle layer, with a confidence interval of 0.272–0.579, p < 0.001.

and conclusion: According to the results presented, three-dimensional anorectal ultra- sonography proved to be a good diagnostic test in the evaluation of the middle and posterior compartments of deep lesions of endometriosis, and there was a correlation between the NMR findings of the pelvis with intestinal preparation in relation to injuries that invade the muscularis propria of the rectum. © 2020 Sociedade Brasileira de Coloproctologia. Published by Elsevier Editora Ltda. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/). ∗ Corresponding author. E-mail: murilo [email protected] (M.M. Tomiyoshi). https://doi.org/10.1016/j.jcol.2020.05.015 2237-9363/© 2020 Sociedade Brasileira de Coloproctologia. Published by Elsevier Editora Ltda. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Article published online: 2021-03-08 244 j coloproctol (rio j). 2 0 2 0; 4 0(3) :243–246 Correlac¸ão entre achados de ressonância magnética nuclear e ultrassonografia anorretal 3D em pacientes com suspeita de endometriose profunda Palavras-chave: Endometriose Ressonância nuclear magnética Ultrassonografia anorretal 3D Endometriose intestinal r e s u m o Introduc¸ão: A endometriose ’e definida como glaˆndulas endometriais e estroma que ocorrem fora da cavidade uterina. Embora na˜o maligno, tecido endometrial ect ´opico e a inflamac¸a˜o resultante podem causar dismenorreia, dispareunia, dor croˆnica e infertilidade. Exames de diagnóstico por imagem mais utilizados são Ressonância Nuclear Magnética e ultrassono- grafia. Métodos: Correlacionar os achados da ultrassonografia anorretal tridimensional com os achados da ressonaˆncia magn´etica nuclear de pelve com preparo intestinal em mulheres portadoras de endometriose profunda, através de um estudo retrospectivo, observacional, transversal, avaliou 63 pacientes do sexo feminino com suspeita de endometriose profunda com prov´avel acometimento intestinal. O per´iodo de avaliac¸ão foi Marc¸o de 2016 a Abril de 2018. Foi realizada análise estatística por meio do Coeficiente de Concordância de Kappa para se avaliar o grau de concordância entre RNM × USG 3D em relac¸ão ao grau de infiltrac¸ão na camada muscular retal com cálculo de intervalo de confianc¸a de 0,272–0,579; p < 0,001. Resultados e conclusão: Segundo os resultados apresentados a ultrassonografia anorretal tridimensional mostrou-se um bom exame diagnóstico na avaliac¸ão dos compartimentos médio e posterior de lesões profundas de endometriose e há correlac¸ão entre os achados da ressonaˆncia magn´etica nuclear de pelve com preparo intestinal em relac¸ão a lesões que invadem a muscular própria do reto. © 2020 Sociedade Brasileira de Coloproctologia. Publicado por Elsevier Editora Ltda. Este ´e um artigo Open Access sob uma licenc¸a CC BY-NC-ND (http://creativecommons.org/ licenses/by-nc-nd/4.0/).

Endometriosis is defined as the presence of endometrial glands and stroma outside the uterine cavity. They are usu- ally located in the pelvis, but can occur in several places, including the intestine, diaphragm, and pleural cavity. Studies report that between 15% and 30% of women with endometrio- sis will have a profound infiltrative disease,1–3 and pregnancy can often be difficult and challenging. One study reported an incidence of intestinal involvement of 3%–37%, 1 while other authors reported an incidence of up to 80%.4 The most common areas of intestinal involvement are the rectum and sigmoid colon. When the gastrointestinal tract is involved, the foci are more commonly located in the sigmoid and rectum, starting from the uterosacral ligament and/or the rectovaginal septum.1 The imaging exam must be able to indicate the num- ber of foci present, the size and depth of the lesion, as well as its distance from the anal margin.5,6 With this information, it is possible to determine the most efficient surgical procedure for each case. Currently, the main role of nuclear magnetic resonance imaging (NMR) in deep pelvic endometriosis lies in the diagnosis and thorough lesion mapping.7 Anorectal ultra- sonography has high sensitivity and specificity (97%–100% and 97%–100%, respectively) for the diagnosis of rectal involvement in patients with endometriosis.8 It also allows measuring the distance from the endometriotic focus to the sphincter apparatus, suggesting which anastomosis will be the most appropriate in each case, as well as the possible need for a stoma. Therefore, it is possible to plan the appropriate surgical approach for each case.9

Correlate the findings of three-dimensional anorectal ultra- sound (USG) with the findings of pelvic NMR with intestinal preparation in women with deep endometriosis.

This was a retrospective, observational, cross-sectional study that evaluated 63 female patients with suspected deep endometriosis and probable intestinal involvement. The eval- uation period was from March 2016 to April 2018. Patients were selected at gynecology service that is a reference in the screening of patients with deep endometriosis for colorectal surgery. The patients were initially submitted to three-dimensional anorectal USG with the BK device with three-dimensional acquisitions. After fasting for six hours, patients underwent a rectal enema with 125 mL of sodium phosphate two hours before the exam and another one hour before the ultrasound. The patients were placed in lateral decubitus (Sims’ position) and underwent anesthetic sedation performed by an anes- thesiologist with propofol and 1 mL of fentanyl. The following variables were analyzed in 3D rectal USG: lesions in perirectal fat, invasion of the muscularis propria, and distance from the lesion to the sphincter apparatus. j coloproctol (rio j). 2 0 2 0; 4 0(3) :243–246 245 The resonance was performed by a radiologist trained in the assessment of patients with deep endometriosis with a high field device (1.5 T) with multiplanar slices of high spatial resolution for fat-suppressed T1- and T2-weighted imaging. All patients underwent intestinal preparation, and vaginal and rectal contrasts were used. The NMR assessed the presence of lesions in the bladder, cervix, uterus, ovaries, round and uterosacral ligaments, rectum, and sigmoid. Patients who did not undergo one of the exams, those in whom cancer was detected at any time during the evaluation, or those who did not have a properly completed medical record were excluded. An Excel spreadsheet was used for data analysis. The data evaluated on NMR were the presence of lesions in the blad- der, cervix, right ovary, left ovary, round ligament, uterosacral ligament, and uterus, and description of the location and size of the lesion in sigmoid and rectum. In 3D anorectal USG, the presence of the size of the lesion, the location of the lesion and whether it invaded the muscularis propria, and the distance from the lesion of the sphincter were assessed. Statistical analysis was performed using the kappa agreement coefficient to assess the degree of agreement between NMR and 3D USG regarding the degree of infil- tration in the rectal muscle layer with a confidence interval of 0.272–0.579 (p < 0.001). The calculation was per- formed using an agreement analysis device available at www.lee.dante.com.br/pesquisa.html.

NMR indicated that six (9.5%) patients had lesions suggestive of bladder endometriosis, one (1.5%) patient had a lesion in the cervix, eight (12%) patients had endometrioma in the left ovary, 11 (17%) patients had endometrioma in the right ovary, eight (12%) patients had lesions in the round ligament, 11 (17%) patients had lesions in the uterosacral ligament, 25 (39%) had lesions suggestive of adenomyosis, 30 (47%) patients had rectal lesions, and 16 (25%) had sigmoid lesions. The analysis of 3D anorectal USG data indicated that nine (14.28%) patients had no lesions, 20 (31.74%) patients had lesions of the perirectal fat, and 28 (44.44%) patients had lesions that invaded the muscularis propria. In this analysis, six (9.52%) patients were excluded from the study due to lack of data. When comparing NMR with 3D USG, it was observed that NMR diagnosed 20 (31%) patients without lesion suggestive of endometriosis, while anorectal USG identified only nine (14%) patients. When comparing the presence of lesion until the perirectal fat, USG identified 20 (31%) patients, while NMR identified 12 (19%) patients. In the comparison between lesions that invaded the muscularis propria, 3D USG diag- nosed 28 (44%), while NMR diagnosed 30 (47%) patients. Statistical analysis was performed using Kappa’s Agree- ment Coefficient to assess the degree of agreement between NMR and 3D USG regarding the degree of infiltration in the rectal muscle layer. The kappa value can be interpreted as follows: <0 — no agreement; 0.0–0.19 — poor; 0.20–0.39 — regular; 0.40–0.59 — moderate; 0.60–0.79 — substantial; 0.80–1.0 — near perfect agreement In patients without lesions, the kappa value was 0.349 (95% CI 0.542–0.155); in the evaluation of lesions in perirectal fat, the kappa value was 0.064 (95% CI: 0.119 to −0.248) and, in lesions that invaded the muscularis propria, the kappa value was 0.864 (95% CI 1–0.611).

Endometriosis is characterized by the presence of endome- trial tissue outside the uterine cavity. The theory of retrograde menstruation is accepted as the most likely etiology, but there appears to be other etiological associations, such as molec- ular biology.10 It affects 10%–15% of women of childbearing age. Intestinal involvement occurs in 3%–7% of these women, with the rectum, sigmoid, or both affected in 90% of these cases.11 The clinical picture is varied; the main complaints, when present, are dyspareunia, dysmenorrhea, chronic pelvic pain, and infertility.10 The definitive diagnosis is made through a surgical procedure with resection of the lesion and subsequent anatomopathological analysis.10 Some studies indicate pelvic ultrasound as a first-line diagnostic test, leaving NMR and endoscopic USG of the rectosigmoid as second and third lines of diagnosis,12 although some systematic reviews demon- strated that both NMR and transvaginal pelvic USG have similar sensitivity and specificity in the diagnosis of deep endometriosis.13 Some studies14 compared the accuracy of two-dimensional vs. three-dimensional USG diagnosis and showed no sig- nificant difference between positive and negative predictive values; however, this comparison was made with the use of transvaginal USG rather than rectal USG.12 Surgical treatment of deep endometriosis involves surgical techniques known as muscle layer dissection (shaving), disc resection, and intestinal resection (segmental or patch). The advantages of the first and second techniques include a pro- cedure with shorter operative time, shorter hospital stay, and fewer complications. In turn, the resection technique can lead to longer surgical time and hospital stay, paralytic ileum, and intestinal complications such as fistulas. The choice of the sur- gical technique depends on the number of lesions observed in imaging exams, their location, and the surgeon’s experience; however, the shaving technique is chosen whenever possible, especially in cases of single and small lesions without invasion of the submucosa.12 Recent systematic reviews indicated that the sensitivity of transvaginal USG ranges from 73.3% to 98.1%, while the NMR sensitivity varies from 73.3% to 100%. The specificity of transvaginal USG ranges from 66.7% to 100%, while that of NMR ranges from 50% to 100%.13 In the present study, 3D anorectal USG had almost perfect agreement, according to the kappa coefficient, when comparing lesions that invade the muscularis propria; however, such expressive agreement val- ues were not observed in patients without lesions or those with lesions in the perirectal fat. In this service, 3D anorec- tal USG showed a significant correlation in the diagnosis of deep intestinal pelvic endometriosis and involvement of the muscularis propria. 246 j coloproctol (rio j). 2 0 2 0; 4 0(3) :243–246 Despite the small number of studies retrieved in the inter- national literature, 3D anorectal USG, despite requiring further studies, can be used as a parameter of choice for surgical treat- ment, and can be used (in combination with the experience of the service/surgeon and/or associated with other imaging exams such as NMR) to define the best surgical technique (shaving or segmental resection).12

3D anorectal USG is a good diagnostic test for the evaluation of the middle and posterior compartments of deep endometrio- sis lesions; a correlation was observed between the findings of NMR of the pelvis with intestinal preparation in relation to injuries that invade the muscularis propria. Conflicts of interest The authors declare no conflicts of interest. r e f e r e n c e s 1. Roseau G, Dumontier I, Palazzo L, Chapron C, Dousset B, Chaussade C, et al. Rectosigmoid endometriosis: endoscopic ultrasound features and clinical implications. Endoscopy. 2000;32:525–30. 2. Keckstein J, Ulrich U, Kandolf O, Wiesinger H, Wustlich M. Laparoscopic therapy of intestinal endometriosis and the ranking of drug treatment. Zentralbl Gynakol. 2003;125:259–66. 3. Fauconnier A, Chapron C. Endometriosis and pelvic pain: epidemiological evidence of the relationship and implications. Hum Reprod Update. 2005;11:595–606. 4. Mezzi G, Ferrari S, Arcidiacono PG, Di Puppo F, Candiani M, Testoni PA. Endoscopic rectal ultrasound and elastosonography are useful in flow chart for the diagnosis of deep pelvic endometriosis with rectal involvement. J Obstet Gynaecol Res. 2011;37:586–90. 5. Chapron C, Vieira M, Chopin N, Balleyguier C, Barakat H, Dumontier I, et al. Accuracy of rectal endoscopic ultrasonography and magnetic resonance imaging in the diagnosis of rectal involvement for patients presenting with deeply infiltrating endometriosis. Ultrassound Obstet Gynecol. 2004;24:175–9. 6. Gonc¸alves MO, Dias JA, Podgaec S Jr, Averbac M, Abrão MS. Transvaginal ultrasound for diagnosis of deeply infiltrating endometriosis. Int J Gynaecol Obstet. 2009;104:156–60. 7. Roy C, Balzan C, Thoma V, Sauer B, Wattiez A, Leroy J. Efficiency of MR imaging to orientate surgical treatment of posterior deep pelvic endometriosis. Abdom Imaging. 2009;34:251–9. 8. Doniec JM, Kahlke V, Peetz F, Schniewind B, Mundhenke C, Lohnert MS, et al. Rectal endometriosis: high sensitivity and specificity of endorectal ultrasound with an impact for the operative management. Dis Colon Rectum. 2003;46:1667–73. 9. Schröder J, Löhnert M, Doniec JM, Dohrmann P. Endoluminal ultrasound diagnosis and operative management of rectal endometriosis. Dis Colon Rectum. 1997;40:614–7. 10. Nácul AP, Spritzer PM. Aspectos atuais do diagnóstico e tratamento da endometriose. Rev Bras Ginecol Obstet. 2010;32:298–307. 11. Gomes NA, Alves K. A ressonância magnética no diagnóstico de endometriose profunda com acometimento intestinal: relato de caso. Unilus Ensino Pesquisa. 2018;38:25–39. 12. Desplats V, Vitte R-L, Cheyron J, Roseau G, Fauconnier A, Moryoussef F. Preoperative rectosigmoid endoscopic ultrasonography predicts the need for bowel resection in endometriosis. World J Gastroenterol. 2019;25:696–706. 13. Moura APC, Ribeiro HSAA, Bernardo WM, Simões R, Torres US, D’Ippolito G, et al. Accuracy of transvaginal sonography versus magnetic resonance imaging in the diagnosis of rectosigmoid endometriosis: systematic review and meta-analysis. PLoS One. 2019;14:e0214842. 14. Guerriero S, Ajossa S, Gerada M, Virgilio B, Angioni S, Melis GB. Diagnostic value of transvaginal ‘tenderne.ss-guided’ ultrasonography for the prediction of location of deep endometriosis. Hum Reprod. 2008;23:2452–7.