Magnetic resonance enema vs rectal water-contrast transvaginal sonography in diagnosis of rectosigmoid

Transcription

1 Ultrasound Obstet Gynecol 2017; 49: Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: /uog Magnetic resonance enema vs rectal water-contrast transvaginal sonography in diagnosis of rectosigmoid endometriosis U. LEONE ROBERTI MAGGIORE*, E. BISCALDI, V. G. VELLONE, P. L. VENTURINI* and S. FERRERO* *Academic Unit of Obstetrics and Gynaecology, IRCCS AOU San Martino IST, Genoa, Italy; Department of Neurosciences, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DiNOGMI), University of Genoa, Genoa, Italy; Department of Radiology, Galliera Hospital, Genoa, Italy; Department of Surgical and Diagnostic Sciences, IRCCS San Martino Hospital and National Institute for Cancer Research, University of Genoa, Genoa, Italy KEYWORDS: bowel endometriosis; diagnosis; endometriosis; magnetic resonance; transvaginal ultrasonography ABSTRACT Objective To compare the accuracy of magnetic resonance enema (MR-e) and rectal water-contrast transvaginal sonography (RWC-TVS) in the diagnosis of rectosigmoid endometriosis. Methods This prospective study included 286 patients of reproductive age with clinical suspicion of rectosigmoid endometriosis. Patients underwent MR-e and RWC-TVS before laparoscopic excision of endometriotic lesions. The findings of MR-e and RWC-TVS were compared with surgical and histological results. Results Of the 286 patients included in the study, 151 (52.8%) had rectosigmoid endometriosis. MR-e and RWC-TVS had similar accuracy in the diagnosis of rectosigmoid endometriosis (P = 0.063). In the diagnosis of rectosigmoid endometriosis with MR-e, the sensitivity was 95.4% (95% CI, %), specificity was 97.8% (95% CI, %), positive predictive value (PPV) was 98.0% (95% CI, %), negative predictive value (NPV) was 95.0% (95% CI, %), positive likelihood ratio (LR+) was (95% CI, ) and negative likelihood ratio (LR ) was 0.05 (95% CI, ). For diagnosis with RWC-TVS, sensitivity was 92.7% (95% CI, %), specificity was 97.0% (95% CI, %), PPV was 97.2% (95% CI, %), NPV was 92.3% (95% CI, %), LR+ was (95% CI, ) and LR was 0.08 (95% CI, ). MR-e and RWC-TVS underestimated the size of the endometriotic nodules; for both imaging techniques the underestimation was greater for nodules with a diameter 30 mm. There was no significant difference in the mean intensity of pain experienced by the patients during the two examinations. Conclusions RWC-TVS should be the first-line investigation in patients with clinical suspicion of rectosigmoid endometriosis and physicians should be trained in performing this examination. Considering that MR-e is more expensive than RWC-TVS, it should be used only when the findings of RWC-TVS are unclear. Copyright 2016 ISUOG. Published by John Wiley & Sons Ltd. INTRODUCTION Rectosigmoid endometriosis is a chronic and progressive benign disease characterized by infiltration of the intestinal wall to at least the muscularis propria, which may cause severe pain, gastrointestinal symptoms, infertility and, in some patients, bowel occlusion 1. Early diagnosis and adequate therapy are mandatory in the management of patients affected by this condition. Medical therapies (such as progestins, combined estroprogestin contraceptives and gonadotropin-releasing hormone analogs) may be administered to treat symptoms caused by rectosigmoid endometriotic nodules that do not cause subocclusive symptoms 2 4. Surgery, intestinal shaving 5,6 or segmental resection 7 should be considered according to the characteristics of the nodules (size and depth of infiltration, degree of stenosis of the bowel lumen) and preference of the patient (desire to conceive or to avoid hormonal therapy). Assessment of the presence and characteristics of rectosigmoid endometriotic nodules is relevant to providing the patient with appropriate and exhaustive coun- Correspondence to: Dr S. Ferrero, Academic Unit of Obstetrics and Gynaecology, IRCCS AOU San Martino IST, Largo R. Benzi 10, Genoa, Italy ( simone.ferrero@unige.it) Accepted: 5 April 2016 ORIGINAL PAPER

2 Diagnosis of rectosigmoid endometriosis by MR-e and RWC-TVS 525 seling, to planning the type of surgery and to scheduling the assistance of a colorectal surgeon. Until recently, the sonographic diagnosis of endometriosis was limited to patients with ovarian endometriomas, and other imaging techniques were used for the assessment of rectosigmoid endometriosis, including rectal endoscopic sonography 8, double-contrast barium enema 9, magnetic resonance imaging (MRI) 10, multidetector computerized tomography enema 11 and virtual colonoscopy 12. However, two recent meta-analyses showed that rectosigmoid endometriosis can be diagnosed accurately using transvaginal sonography (TVS) 13,14. Rectosigmoid distension may facilitate the identification of intestinal lesions during both MRI and TVS 18,19. However, no previous study compared the accuracy of magnetic resonance enema (MR-e) and rectal water-contrast TVS (RWC-TVS) in the diagnosis of rectosigmoid endometriosis. The objective of this prospective study was to compare the accuracy of MR-e and RWC-TVS in diagnosing rectosigmoid endometriotic nodules. SUBJECTS AND METHODS This prospective study included all consecutive patients who were referred to our institution because of clinical suspicion of rectosigmoid endometriosis between November 2008 and December During this period, the imaging workup required that both MR-e and RWC-TVS were performed in patients with suspicion of rectosigmoid endometriosis. The local ethics committee approved the study protocol. Patients participating in the study provided written informed consent. Inclusion criteria for the study were the following: reproductive age and suspicion of deep pelvic endometriosis on the basis of gynecological symptoms and vaginal examination and/or presence of gastrointestinal symptoms that might be caused by rectosigmoid endometriosis. The following exclusion criteria were used: previous bilateral ovariectomy, previous examinations diagnosing bowel endometriosis (such as double-contrast barium enema, rectal endoscopic sonography or multidetector computed tomography enema), previous bowel surgery (except appendectomy), renal or hepatic failure or presence of contraindications to MR examination. Symptoms were investigated systematically during the study period and were recorded in a database. The presence of dysmenorrhea, deep dyspareunia, chronic pelvic pain and dyschezia was assessed; the intensity of these symptoms was rated on a 100-mm visual analog scale (VAS), the extreme left of the scale representing absence of pain and the extreme right of the scale indicating maximal intensity of pain. The presence of the following gastrointestinal symptoms was assessed: diarrhea (more than three bowel movements per day), constipation during the menstrual cycle, abdominal bloating, intestinal cramping and feeling of incomplete evacuation after bowel movements. A symptom analog scale questionnaire was used to estimate the severity of each gastrointestinal symptom (1 indicated absence of the symptom and 10 indicated highest severity of the symptom). Two physicians performed MR-e and RWC-TVS independently. The radiologist (E.B.) and the sonographer (S.F.) knew the clinical data and that rectosigmoid endometriosis was suspected; however, each was blinded to the findings of the other imaging technique. All patients included in the study underwent laparoscopy within 3 months from the completion of diagnostic investigations. Intestinal endometriosis was defined as the disease infiltrating at least into the muscularis propria. Endometriotic foci located on the bowel serosa were considered peritoneal and not bowel endometriosis. The decision to perform segmental bowel resection was based on the intraoperative findings. The findings of MR-e and RWC-TVS were compared with surgical and histological results, which were considered the gold standard. The primary objective of the study was to assess the accuracy of the two imaging techniques in diagnosing the presence of rectosigmoid endometriosis. The secondary objective was to assess the accuracy of the two imaging techniques in diagnosing the presence of infiltration of the mucosal layer of the bowel wall. The tertiary objective was to compare the precision of the two techniques in estimating the size of the rectosigmoid endometriotic nodules. Magnetic resonance enema A radiologist (E.B.) performed all the exams according to a standardized procedure 17. MR-e was performed on a 1.5-T magnet (Signa Excite HDx, GE Medical Systems, Zipf, Austria) using an 8-channel phased-array coil. The position of the patient was preferably horizontal and the entry position in the MR gantry was feet first. Pelvic volume was evaluated. MR-e studies followed a standardized protocol consisting of the following: T2-weighted (T2W) fast-recovery fast spin-echo (FrFSE) axial images: field of view (FOV) = cm; echo time (TE) = 150 ms; repetition time (TR) = 5500 ms; matrix = ; slice thickness = 3 mm; spacing = 1; acceleration factor = 1; number of excitations (NEX) = 4; T2W FrFSE fat saturation coronal images: FOV = cm; TE = 150 ms; TR = 4250 ms; matrix = ; slice thickness = 4 mm; spacing = 1; acceleration factor = 1; NEX = 4; T1-weighted (T1W) FSE coronal images: FOV = cm; TE = min full; TR = 550 ms; matrix = ; slice thickness = 4 mm; spacing = 1; acceleration factor = 1; NEX = 2; fast imaging employing steady-state acquistion sequence in coronal plane: FOV = 37 cm; slice thickness = 3 mm; spacing = 1; matrix = ; NEX = 2; T2W FrFSE sagittal images: FOV = cm; TE = 150 ms; TR = 4250 ms; matrix = ; slice thickness = 3 mm; spacing = 1; acceleration factor = 1; NEX = 6; T2W FrFSE fat saturation sagittal images: FOV = cm; TE = 120 ms; TR= 4600 ms; matrix = ; slice thickness = 4 mm; spacing = 1; acceleration factor = 1.5; NEX = 4; diffusion-weighted echo-planar imaging (b = 800) Ultrasound Obstet Gynecol 2017; 49:

3 Leone Roberti Maggiore et al. 526 U U U R B R R B C R C C Figure 1 Magnetic resonance enema (MR-e) T1-weighted (T1W) and T2-weighted (T2W) sagittal images of rectal endometriotic nodule. (a) Fast-recovery fast spin-echo (FSE) T2W sequence: nodule has T2 hyperintense spots. (b) FSE T1W fat saturation sequence: nodule has hemorrhagic focal hyperintense spots. (c) FSE T1W fat saturation sequence after gadolinium injection: nodule shows generic enhancement similar to that of uterine and rectal walls. Endometriotic nodule is seen infiltrating muscularis mucosae of middle rectum (arrowheads). B, bladder; C, catheter used to dilate rectosigmoid colon; R, rectum; U, uterus. axial images: FOV = cm; TR = 3000 ms; matrix = ; slice thickness = 4 mm; spacing = 1; NEX = 8. T1W images were acquired, employing fat-suppression even after contrast enhancement (gadobutrol at a dosage of 0.2 mmol/kg body weight; Gadovist 1.0, Schering, Berlin, Germany) (Figure 1). After the patient lay on the MR bed, retrograde distension was performed initially in the left lateral decubitus, then in the horizontal position to reduce abdominal wall movements and respiratory artifacts. Sonographic gel (Aquasonic, Parker Laboratories, Fairfield, NJ, USA), ml, diluted with saline solution (1 : 8), was introduced using a syringe connected to a 20-Fr Foley catheter to distend the rectum and the sigmoid colon. Intestinal hypotonization was not used. Endometriotic nodules appear as solid masses outside the rectosigmoid wall, frequently with a hypointense signal due to their fibrous nature10,20. Visible penetration of the nodules into the intestinal wall was the main parameter to diagnose infiltration of the muscularis propria17 (Figure 2). Rectal water-contrast transvaginal sonography A few hours before sonography, a rectal enema (133 ml of monobasic sodium phosphate anhydrous; Clisma Lax; Sofar, Milan, Italy) was used to clean the rectosigmoid colon of fecal residue. The examinations were assessed in real time and archived, with images saved as TIFF files. A physician (S.F.) performed the exams according to a standardized protocol using a Voluson E6 ultrasound machine (GE Medical Systems). After the transducer was introduced into the vagina, an assistant inserted a 6-mm (18 Ch) flexible catheter through the anal os into the rectal lumen up to 15 cm from the anus. A gel, infused with lidocaine (Luan, Molteni & C., Scandicci, Italy), was used to minimize the discomfort due to the passage of the catheter. After the connection of a 50-mL syringe to the catheter, warm sterile saline solution was injected into the rectosigmoid under sonographic guidance. One hundred milliliters of saline solution was infused continuously at the beginning of the procedure; subsequently, additional saline solution (up to 300 ml) was injected if requested by the sonographer, depending on the distensibility of the intestinal wall. During the examination, Klemmer forceps were placed on the catheter to prevent backflow of the saline solution through the catheter when the solution was not being injected. There was no significant leakage of saline solution into the space between the catheter and the anus. Images were obtained before, during and after saline injection. On ultrasound, intestinal serosa is hyperechoic, the two layers of the muscularis propria appear as hypoechoic strips separated by a fine hyperechoic line, the submucosa is hyperechoic, the muscularis mucosa is hypoechoic and the interface between the lumen and the mucosal layer is hyperechoic21. Rectosigmoid endometriotic nodules appear as nodular, solid, hypoechoic lesions, adjacent to and/or penetrating the muscularis propria; these nodules usually cause a thickening of the intestinal wall (Figure 3). Hyperechoic foci may sometimes be present within the nodule. Retraction and adhesions are often present; this results in the so-called Indian headdress or comet sign22,23. Distension of the intestinal wall with saline solution facilitated the identification of the limits of the intestinal nodules. Tolerability of examinations Patients were asked to rate the pain experienced during MR-e and RWC-TVS using a 100-mm VAS; the extreme left represented the absence of pain, and the extreme Ultrasound Obstet Gynecol 2017; 49:

4 Diagnosis of rectosigmoid endometriosis by MR-e and RWC-TVS 527 right represented the worst possible pain. Mild pain was defined as VAS score < 2, moderate pain as VAS score between 2 and 5 and severe pain as VAS score > 5. Surgical technique Before laparoscopy, the surgeons examined the reports and the images from both MR-e and RWC-TVS. A team of gynecological and colorectal surgeons with extensive experience in the surgical treatment of pelvic and rectosigmoid endometriosis performed all the procedures laparoscopically. After adequate adhesiolysis, the sigmoid colon and the rectum were examined systematically to identify the presence of endometriotic lesions. All visible endometriotic nodules (except those on the diaphragm) were excised. Segmental bowel resection was performed when rectosigmoid endometriotic lesions infiltrated at least the muscularis propria. The landmark used to define the transition from the sigmoid colon to the rectum was the loss of the teniae coli, the appendices epiploicae and the surgical mesocolon that is located anatomically at about the level of the third sacral vertebra. In particular, we defined the level at which gradual convergence of the teniae coli started as the rectosigmoid junction, while the point at which teniae coli could no longer be seen was considered as the beginning of the upper rectum. The surgical specimens were evaluated histologically; the depth of infiltration of the endometriotic nodules in the bowel wall was assessed24. In addition, the presence of multifocal disease (presence of one or more lesions that affected the sigmoid colon and that were associated with the colorectal primary lesion) was assessed. Statistical analysis 500 μm Figure 2 Rectal endometriotic nodule on magnetic resonance enema (MR-e) (a,b) and histological examination of resected nodule (c). MR-e fast-recovery fast spin-echo T2 (a) and T2-weighted fat saturation sequence (b) in sagittal plane, showing rectal nodule deeply infiltrating anterior wall of lower rectum. Nodule is inhomogeneously hypointense in both sequences. Liquid instilled in rectum enhances lesion detectability. Fat saturation confirms space between rectum and uterus is infiltrated. It is possible to detect two layers peripheral to nodule: deeper layer shows considerable infiltration of muscular wall (arrowheads) and layer on surface of nodule (arrow) shows rectal mucosa is not infiltrated. (c) Hematoxylin and eosin-stained section of resected rectal nodule confirms rectal mucosa is not infiltrated by endometriosis. Accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+) and negative likelihood ratio (LR ) were calculated for both MR-e and RWC-TVS. In cases of multifocal disease, the main nodule affecting the bowel was considered for the analysis. McNemar s test with Yates s continuity correction was used to compare the accuracy of MR-e and RWC-TVS in the diagnosis of rectosigmoid endometriosis. The precision of the measurement of nodule size by the imaging techniques was estimated by subtracting the size of the nodule as measured by the imaging techniques from the size of the nodule as measured at histology. The non-parametric Mann Whitney U-test was used to compare the intensity of pain experienced by the patients during MR-e and RWC-TVS. The χ2 test was used to compare the type of pain (mild, moderate and severe) experienced by the patients undergoing the two exams. Data were analyzed using the SPSS Statistics version 20.0 (IBM Corp., Armonk, NY, USA). P < 0.05 was considered statistically significant. RESULTS A total of 286 women who underwent both diagnostic examinations and surgery were included in the study. The main characteristics of the study population are Ultrasound Obstet Gynecol 2017; 49:

5 528 Leone Roberti Maggiore et al. Patients with suspicion of rectosigmoid endometriosis (n = 286) MR-e (n = 286) RWC-TVS (n = 286) WC Laparoscopy (n = 286) No surgery (n = 135) Segmental bowel resection (n = 151) Figure 3 Rectal water-contrast (WC) transvaginal sonography of rectal endometriotic nodule ( ). Endometriotic nodule penetrates muscularis propria of rectum, causing thickening of intestinal wall. Normal layers of rectal wall can be observed distal to nodule (arrow). Table 1 Characteristics of 286 women with suspected rectosigmoid endometriosis Characteristic Value Age (years) 31.9 ± 4.8 BMI (kg/m 2 ) 23.6 ± 2.2 Smoker 91 (31.8) Level of education Primary school 7 (2.4) Secondary school 204 (71.3) University 75 (26.2) Previous live birth 68 (23.8) Previous surgery for endometriosis 126 (44.1) Hormonal therapy None 165 (57.7) Sequential oral contraceptive 57 (19.9) Continuous oral contraceptive 18 (6.3) Contraceptive vaginal ring 18 (6.3) Norethisterone acetate 28 (9.8) Data are given as mean ± SD or n (%). BMI, body mass index. summarized in Table 1. The path of the patients through the study is shown in Figure 4. The intensity of pain and gastrointestinal symptoms are presented in Table 2. Surgery and histological examination demonstrated that 151 (52.8% (95% CI, %)) women had rectosigmoid endometriotic nodules. The endometriotic nodules were located on the sigmoid colon in 90 (59.6%) patients, on the rectosigmoid junction in 22 (14.6%) and on the rectum in 39 (25.8%). Multifocal disease was found in 21 (13.9%) patients who had two endometriotic nodules affecting the bowel. The mean ± SD length of the resected bowel segment was 12.5 ± 2.8 cm. The diagnosis of endometriosis was confirmed in all the excised nodules by histological examination. The deeper nodule infiltrated the intestinal muscularis propria in 107 (70.9%) patients, the submucosa in 31 (20.5%) and the mucosa in 13 (8.6%). Histological confirmation of bowel endometriosis (n = 151) Figure 4 Flowchart of patients with suspected rectosigmoid endometriosis through study. MR-e, magnetic resonance enema; RWC-TVS, rectal water-contrast transvaginal sonography. Table 2 Intensity of pain and gastrointestinal symptoms in 286 women with suspected rectosigmoid endometriosis Symptom n (%) Severity (mean ± SD) Dysmenorrhea 244 (85.3) 63.4 ± 17.0* Non-menstrual pelvic pain 235 (82.2) 55.0 ± 13.4* Dyspareunia 230 (80.4) 58.9 ± 15.8* Dyschezia 166 (58.0) 52.5 ± 17.0* Persistent constipation 106 (37.1) 5.4 ± 2.2 Constipation during menstruation 57 (19.9) 6.0 ± 1.6 Diarrhea 80 (28.0) 7.6 ± 1.3 Diarrhea during menstruation 94 (32.9) 8.0 ± 0.9 Intestinal cramping 180 (62.9) 6.4 ± 1.6 Abdominal bloating 169 (59.1) 6.7 ± 1.7 Feeling of incomplete evacuation 105 (36.7) 6.3 ± 1.4 Passage of mucus 103 (36.0) 5.8 ± 1.8 Cyclical rectal bleeding 46 (16.1) 5.0 ± 1.0 *Measured using 100-mm visual analog scale. Measured using 10-point symptom analog scale questionnaire. Accuracy of magnetic resonance enema and rectal water-contrast transvaginal sonography The performance of MR-e and RWC-TVS in the diagnosis of rectosigmoid endometriosis is summarized in Table 3. The two techniques had similar accuracy (96.5% vs 94.8%; P = 0.063) in the diagnosis of rectosigmoid endometriosis (Figures 5 and 6). The performance of MR-e and RWC-TVS in the diagnosis of mucosal infiltration of the endometriosis is summarized in Table 4. The accuracy of RWC-TVS was superior to that of MR-e (86.4% vs 84.3%; P < 0.001) in the detection of infiltration of the mucosal layer. Ultrasound Obstet Gynecol 2017; 49:

6 Diagnosis of rectosigmoid endometriosis by MR-e and RWC-TVS 529 Table 3 Performance of magnetic resonance enema (MR-e) and rectal water-contrast transvaginal sonography (RWC-TVS) in diagnosis of rectosigmoid endometriosis in 286 women Imaging technique MR-e RWC-TVS Accuracy (%) Sensitivity (%) Specificity (%) PPV (%) NPV (%) LR+ LR 96.5 ( ) 94.8 ( ) 95.4 ( ) 92.7 ( ) 97.8 ( ) 97.0 ( ) 98.0 ( ) 97.2 ( ) 95.0 ( ) 92.3 ( ) ( ) ( ) 0.05 ( ) 0.08 ( ) Values in parentheses are 95% CI. LR, negative likelihood ratio; LR+, positive likelihood ratio; NPV, negative predictive value; PPV, positive predictive value. WC 2 mm Figure 5 (a) Rectal water-contrast transvaginal sonography showing hypoechoic rectosigmoid endeometriotic nodule penetrating thickened rectal wall ( ). Hyperechoic foci can be observed. Water contrast (WC) shows that muscularis mucosa (hypoechoic) and interface between lumen and mucosal layer (hyperechoic) are not infiltrated. (b) Magnetic resonance enema (MR-e) in sagittal plane (fast spin-echo (FSE) T2-weighted (T2W) sequence) after retrograde distension of rectosigmoid, showing endometriotic nodule deeply infiltrating anterior wall of lower rectum (arrows). Posterior wall of uterus is in strict relationship with the nodule. (c) MR-e in axial plane (FSE T2W sequence) showing rectal wall infiltrated by endometriosis; intestinal wall is thickened concentrically (arrowhead). (d) Hematoxylin and eosin-stained section of resected rectal nodule confirming endometriotic cysts within muscular layer of rectal wall; columnar endometrial epithelial cells and endometriotic stroma can be observed. Endometriotic cysts are surrounded by thin fibrous capsule and muscle cells. Intestinal submucosa and mucosa are not infiltrated by endometriosis. Ultrasound Obstet Gynecol 2017; 49:

7 Leone Roberti Maggiore et al. 530 WC Figure 6 Rectal endometriotic nodule on magnetic resonance enema (MR-e) (a,b) and rectal water-contrast (WC) transvaginal sonography (c). (a,b) MR-e fast-recovery fast spin-echo T2 (a) and T2-weighted fat saturation sequence (b) (sagittal plane), showing rectal nodule (arrowhead) infiltrating anterior wall of lower rectum; nodule is inhomogeneously hypointense in both sequences. Liquid instilled in rectum enhances intestinal wall visibility and detection of nodule. Muscular wall is infiltrated. (c) WC facilitates identification of rectal endometriotic nodule ( ) with large diameter of 35 mm. Table 4 Performance of magnetic resonance enema (MR-e) and rectal water-contrast transvaginal sonography (RWC-TVS) in diagnosis of infiltration of mucosal layer of bowel wall in 286 women with suspected rectosigmoid endometriosis Imaging technique MR-e RWC-TVS Accuracy (%) Sensitivity (%) Specificity (%) PPV (%) NPV (%) LR+ LR 84.3 ( ) 86.4 ( ) 66.7 ( ) 76.9 ( ) 85.0 ( ) 86.1 ( ) 16.3 ( ) 20.8 ( ) 98.3 ( ) 98.7 ( ) 4.46 ( ) 5.53 ( ) 0.39 ( ) 0.27 ( ) Values in parentheses are 95% CI. LR, negative likelihood ratio; LR+, positive likelihood ratio; NPV, negative predictive value; PPV, positive predictive value. Both MR-e and RWC-TVS underestimated the size of the endometriotic nodules; the underestimation was greater for nodules with a diameter 30 mm for both imaging techniques (Table 5). The mean ± SD overall occupation time of the scanner room was 45.7 ± 3.9 min for MR-e and the mean ± SD time required to perform RWC-TVS was 18.2 ± 2.9 min (P < 0.001). Tolerability and adverse effects of magnetic resonance enema and rectal water-contrast transvaginal sonography MR-e was performed safely in all patients; there was no adverse reaction to paramagnetic contrast medium. During injection of the sonographic gel into the rectosigmoid, none of the patients showed signs of severe discomfort or had an allergic reaction. RWC-TVS was always well tolerated and no patient requested to interrupt the exam. The mean ± SD intensity of pain experienced during the examinations was similar for MR-e and RWC-TVS (VAS score, 3.6 ± 1.5 vs 3.4 ± 1.2; P = 0.098). Severe pain was experienced by 44 (15.4%) women during MR-e and 27 (9.4%) women during RWC-TVS, moderate pain was experienced by 115 (40.2%) during MR-e and 108 (37.8%) during RWC-TVS, and mild pain was experienced by 127 (44.4%) during MR-e and 151 (52.8%) during RWC-TVS (P = 0.042). Ultrasound Obstet Gynecol 2017; 49:

8 Diagnosis of rectosigmoid endometriosis by MR-e and RWC-TVS 531 Table 5 Difference between size of largest nodule estimated by magnetic resonance enema (MR-e) and rectal water-contrast transvaginal sonography (RWC-TVS), and that measured on histopathology Largest diameter on histology MR-e RWC-TVS (mm, mean ± SD) MD (95% CI) (mm)* LOA (mm) MD (95% CI) (mm)* LOA (mm) All nodules (n = 286) 27.6 ± (1.2 to 1.9) 5.1 to (1.3 to 2.0) 4.9 to 8.2 Nodules with diameter: < 30 mm (n = 198) 23.9 ± (0.6 to 1.2) 3.4 to (0.8 to 1.4) 3.7 to mm (n = 88) 36.1 ± (2.0 to 4.0) 6.5 to (1.9 to 3.8) 6.1 to 11.8 *Mean difference (MD) calculated by subtracting size of nodule measured by imaging technique from size of nodule measured on histological examination. Limits of agreement (LOA) calculated as MD ± 2 SD of the difference. DISCUSSION This is the first study demonstrating that MR-e and RWC-TVS have similar accuracy in the diagnosis of rectosigmoid endometriosis. In both techniques, rectosigmoid distension was used to facilitate the identification of rectosigmoid endometriotic nodules. Over the last decade, several ultrasound techniques have been used to study women with suspected rectosigmoid endometriosis, including tenderness-guided ultrasound 25, sonovaginography with saline or gel and three-dimensional (3D) sonography 29. The choice of the sonographic technique is often based on the experience of the sonographer rather than on evidence of superiority of one technique compared with the others. In this study, RWC-TVS was chosen for the comparison with MR-e because of the personal experience of the authors and of the common criterion of bowel distension with fluid. Previous studies applied intestinal jelly distension to MR for the diagnosis of deep endometriosis 16, Ten years ago, a prospective study showed that the injection of ultrasound jelly in the vagina (20 30 ml) and the rectum (150 ml) during MR facilitated the identification of complete cul-de-sac obliteration in 31 patients with suspected rectovaginal endometriosis 30. The usefulness of this technique was confirmed subsequently by the same authors in larger series 31,33. Furthermore, others confirmed that intestinal distension and opacification using ultrasound gel helps to visualize rectosigmoid endometriotic nodules 16,32 and applied this technique to the 3.0-T MR study of patients with suspected endometriosis 34. A recent study compared the accuracy of MR-e with multidetector computerized tomography enema in estimating the presence of rectosigmoid endometriotic nodules 17.That study showed that both techniques were accurate in the diagnosis of rectosigmoid endometriosis; in particular, MR-e had accuracy, sensitivity, specificity, PPV and NPV of 96.9%, 97.2%, 96.4%, 98.3% and 94.1%, respectively, in line with the results of the current study. Our group was the first to apply the criteria of retrograde bowel distension to TVS 18,19,35 and this technique was studied subsequently by others 36.More recently, 3D sonography combined with intestinal distension demonstrated promising results in the diagnosis of rectosigmoid endometriosis 37. The effectiveness of RWC-TVS in the diagnosis of rectosigmoid endometriosis observed in that study is similar to that reported previously by us 18 and others 36. A potential limitation of this study may be the experience of the radiologist in performing MR-e and of the sonographer in performing RWC-TVS, which may influence the accuracy of these techniques in diagnosing rectosigmoid endometriosis. However, referral to a tertiary center with expertise in endometriosis is mandatory when bowel involvement is suspected to guarantee to the patient the best diagnostic and therapeutic pathway. A further limitation of this study is that the surgeons were aware of the findings of MR-e and RWC-TVS. Although in an ideal prospective study the surgeons should be blinded to the findings of the preoperative investigations, this theoretical study design appears unethical in clinical practice because diagnostic imaging may facilitate the identification of intestinal endometriotic nodules during surgery. Furthermore, the knowledge of the findings of the preoperative investigations may only help the surgeons in identifying endometriotic nodules that were actually present. A strength of the current study is the large sample size. Out of 286 women with the suspicion of rectosigmoid endometriosis, over 50% of the patients actually had bowel nodules. This study demonstrates that both MR-e and RWC-TVS are accurate in the diagnosis of rectosigmoid endometriosis. RWC-TVS has some advantages in comparison with MR-e: it is faster and does not require the use of contrast medium that may potentially cause allergic reactions. Furthermore, TVS is superior to MR in terms of cost-effectiveness 38 and allows assessment of the degree of infiltration of the muscularis propria 18,19,35,36 and estimation of the degree of stenosis of the intestinal lumen 36, and thus it should be considered the first-line technique to diagnose rectosigmoid endometriosis. The use of MR to study all symptomatic women before any form of treatment (medical or surgical) would increase the cost of preoperative assessment without a true increase in diagnostic performance. However, TVS should be performed by highly skilled sonographers. In fact, it has been estimated recently that the learning curve for an accurate diagnosis of deep pelvic endometriosis by TVS requires performance in about 40 cases 39 ;therefore, it may be difficult for sonographers working in small centers to achieve such experience. The main advantage of MR-e could be that, with a retrograde distension of the entire colon, it may provide a complete overview of the whole colon 40. In the current study, the distension was targeted to the rectosigmoid because the aim of the study Ultrasound Obstet Gynecol 2017; 49:

9 532 Leone Roberti Maggiore et al. was comparison with RWC-TVS, and endometriotic lesions of the right colon are beyond the field of view of a transvaginal approach. Furthermore, the whole colon MR-e examination requires more time. A complete colonic distension may be difficult to tolerate for the time required to complete all MR-e sequences, and appropriate intestinal cleansing is required. In the current study, we did not compare the accuracy of RWC-TVS with TVS alone, which was the objective of a previous study 17. In conclusion, this study shows that RWC-TVS and MR-e have similar accuracy in the diagnosis of rectosigmoid endometriosis. The mean intensity of pain perceived during RWC-TVS and MR-e is similar but severe pain is perceived by the patients more frequently during MR-e. However, the methodology of the two imaging techniques was different (in terms of type of catheters used and volume of fluid instilled) and this may have influenced the discomfort perceived by the patients. Given the better cost-effectiveness, in our opinion, TVS should be the first-line investigation for patients with clinical suspicion of rectosigmoid endometriosis. Educational programs should be developed to offer sonographers who are familiar with the general use of TVS the opportunity to improve their skills in the diagnosis of rectosigmoid endometriosis. REFERENCES 1. Remorgida V, Ferrero S, Fulcheri E, Ragni N, Martin DC. Bowel endometriosis: presentation, diagnosis, and treatment. Obstet Gynecol Surv 2007; 62: Ferrero S, Camerini G, Ragni N, Venturini PL, Biscaldi E, Remorgida V. Norethisterone acetate in the treatment of colorectal endometriosis: a pilot study. Hum Reprod 2010; 25: Ferrero S, Camerini G, Ragni N, Menada MV, Venturini PL, Remorgida V. Triptorelin improves intestinal symptoms among patients with colorectal endometriosis. IntJGynaecolObstet2010; 108: Maggiore ULR, Remorgida V, Scala C, Tafi E, Venturini PL, Ferrero S. Desogestrel-only contraceptive pill versus sequential contraceptive vaginal ring in the treatment of rectovaginal endometriosis infiltrating the rectum: a prospective open-label comparative study. Acta Obstet Gynecol Scand 2014; 93: Donnez J, Squifflet J. Complications, pregnancy and recurrence in a prospective series of 500 patients operated on by the shaving technique for deep rectovaginal endometriotic nodules. Hum Reprod 2010; 25: Roman H. Rectal shaving using PlasmaJet in deep endometriosis of the rectum. Fertil Steril 2013; 100: e Darai E, Dubernard G, Coutant C, Frey C, Rouzier R, Ballester M. Randomized trial of laparoscopically assisted versus open colorectal resection for endometriosis: morbidity, symptoms, quality of life, and fertility. Ann Surg 2010; 251: Bazot M, Lafont C, Rouzier R, Roseau G, Thomassin-Naggara I, Darai E. Diagnostic accuracy of physical examination, transvaginal sonography, rectal endoscopic sonography, and magnetic resonance imaging to diagnose deep infiltrating endometriosis. Fertil Steril 2009; 92: Faccioli N, Foti G, Manfredi R, Mainardi P, Spoto E, Ruffo G, Minelli L, Mucelli RP. Evaluation of colonic involvement in endometriosis: double-contrast barium enema vs. magnetic resonance imaging. Abdom Imaging 2010; 35: Bazot M, Darai E, Hourani R, Thomassin I, Cortez A, Uzan S, Buy JN. Deep pelvic endometriosis: MR imaging for diagnosis and prediction of extension of disease. Radiology 2004; 232: Biscaldi E, Ferrero S, Fulcheri E, Ragni N, Remorgida V, Rollandi GA. Multislice CT enteroclysis in the diagnosis of bowel endometriosis. Eur Radiol 2007; 17: van der Wat J, Kaplan MD. Modified virtual colonoscopy: a noninvasive technique for the diagnosis of rectovaginal septum and deep infiltrating pelvic endometriosis. J Minim Invasive Gynecol 2007; 14: Hudelist G, English J, Thomas AE, Tinelli A, Singer CF, Keckstein J. Diagnostic accuracy of transvaginal ultrasound for non-invasive diagnosis of bowel endometriosis: systematic review and meta-analysis. Ultrasound Obstet Gynecol 2011; 37: Guerriero S, Ajossa S, Orozco R, Perniciano M, Jurado M, Melis GB, Alcazar JL. Diagnostic accuracy of transvaginal ultrasound for diagnosis of deep endometriosis in the rectosigmoid: a meta-analysis. Ultrasound Obstet Gynecol 2016; 47: Chassang M, Novellas S, Bloch-Marcotte C, Delotte J, Toullalan O, Bongain A, Chevallier P. Utility of vaginal and rectal contrast medium in MRI for the detection of deep pelvic endometriosis. Eur Radiol 2010; 20: Loubeyre P, Copercini M, Frossard JL, Wenger JM, Petignat P. Pictorial review: rectosigmoid endometriosis on MRI with gel opacification after rectosigmoid colon cleansing. Clin Imaging 2012; 36: Biscaldi E, Ferrero S, Maggiore ULR, Remorgida V, Venturini PL, Rollandi GA. Multidetector computerized tomography enema versus magnetic resonance enema in the diagnosis of rectosigmoid endometriosis. Eur J Radiol 2014; 83: Valenzano Menada M, Fortunato T, Ferrero S, Petrera P, Fulcheri E, Bogliolo S. Massive ovarian edema: a case of over 30 cm ovarian mass in young woman. Minerva Ginecol 2008; 60: Ferrero S, Biscaldi E, Morotti M, Venturini PL, Remorgida V, Rollandi GA, Valenzano Menada M. Multidetector computerized tomography enteroclysis vs. rectal water contrast transvaginal ultrasonography in determining the presence and extent of bowel endometriosis. Ultrasound Obstet Gynecol 2011; 37: Biscaldi E, Ferrero S, Remorgida V, Fulcheri E, Rollandi GA. Rectosigmoid endometriosis with unusual presentation at magnetic resonance imaging. Fertil Steril 2009; 91: Goncalves MO, Podgaec S, Dias JA Jr, Gonzalez M, Abrao MS. Transvaginal ultrasonography with bowel preparation is able to predict the number of lesions and rectosigmoid layers affected in cases of deep endometriosis, defining surgical strategy. Hum Reprod 2010; 25: Guerriero S, Condous G, van den Bosch T, Valentin L, Leone FP, Van Schoubroeck D, Exacoustos C, Installé AJ, Martins WP, Abrao MS, Hudelist G, Bazot M, Alcazar JL, Gonçalves MO, Pascual MA, Ajossa S, Savelli L, Dunham R, Reid S, Menakaya U, Bourne T, Ferrero S, Leon M, Bignardi T, Holland T, Jurkovic D, Benacerraf B, Osuga Y, Somigliana E, Timmerman D. Systematic approach to sonographic evaluation of the pelvis in women with suspected endometriosis, including terms, definitions and measurements: a consensus opinion from the International Deep Endometriosis Analysis (IDEA) group. Ultrasound Obstet Gynecol 2016; 48: Benacerraf BR, Groszmann Y, Hornstein MD, Bromley B. Deep infiltrating endometriosis of the bowel wall: the comet sign. J Ultrasound Med 2015; 34: Valenzano Menada M, Remorgida V, Abbamonte LH, Nicoletti A, Ragni N, Ferrero S. Does transvaginal ultrasonography combined with water-contrast in the rectum aid in the diagnosis of rectovaginal endometriosis infiltrating the bowel? Hum Reprod 2008; 23: Guerriero S, Ajossa S, Gerada M, D Aquila M, Piras B, Melis GB. Tenderness-guided transvaginal ultrasonography: a new method for the detection of deep endometriosis in patients with chronic pelvic pain. Fertil Steril 2007; 88: Saccardi C, Cosmi E, Borghero A, Tregnaghi A, Dessole S, Litta P. Comparison between transvaginal sonography, saline contrast sonovaginography and magnetic resonance imaging in the diagnosis of posterior deep infiltrating endometriosis. Ultrasound Obstet Gynecol 2012; 40: León M, Vaccaro H, Alcázar JL, Martinez J, Gutierrez J, Amor F, Iturra A, Sovino H. Extended transvaginal sonography in deep infiltrating endometriosis: use of bowel preparation and an acoustic window with intravaginal gel: preliminary results. J Ultrasound Med 2014; 33: Reid S, Lu C, Hardy N, Casikar I, Reid G, Cario G, Chou D, Almashat D, Condous G. Office gel sonovaginography for the prediction of posterior deep infiltrating endometriosis: a multicenter prospective observational study. Ultrasound Obstet Gynecol 2014; 44: Guerriero S, Saba L, Ajossa S, Peddes C, Angiolucci M, Perniciano M, Melis GB, Alcázar JL. Three-dimensional ultrasonography in the diagnosis of deep endometriosis. Hum Reprod 2014; 29: Takeuchi H, Kuwatsuru R, Kitade M, Sakurai A, Kikuchi I, Shimanuki H, Kinoshita K. A novel technique using magnetic resonance imaging jelly for evaluation of rectovaginal endometriosis. Fertil Steril 2005; 83: Kikuchi I, Takeuchi H, Kuwatsuru R, Kitade M, Kumakiri J, Kuroda K, Takeda S. Diagnosis of complete cul-de-sac obliteration (CCDSO) by the MRI jelly method. J Magn Reson Imaging 2009; 29: Loubeyre P, Petignat P, Jacob S, Egger JF, Dubuisson JB, Wenger JM. Anatomic distribution of posterior deeply infiltrating endometriosis on MRI after vaginal and rectal gel opacification. Am J Roentgenol 2009; 192: Kikuchi I, Kuwatsuru R, Yamazaki K, Kumakiri J, Aoki Y, Takeda S. Evaluation of the usefulness of the MRI jelly method for diagnosing complete cul-de-sac obliteration. Biomed Res Int 2014; 2014: Hottat N, LarrousseC, AnafV, Noël JC, Matos C, Absil J, Metens T. Endometriosis: contribution of 3.0-T pelvic MR imaging in preoperative assessment initial results. Radiology 2009; 253: Menada MV, Remorgida V, Abbamonte LH, Fulcheri E, Ragni N, Ferrero S. Transvaginal ultrasonography combined with water-contrast in the rectum in the diagnosis of rectovaginal endometriosis infiltrating the bowel. Fertil Steril 2008; 89: Bergamini V, Ghezzi F, Scarperi S, Raffaelli R, Cromi A, Franchi M. Preoperative assessment of intestinal endometriosis: A comparison of transvaginal sonography with water-contrast in the rectum, transrectal sonography, and barium enema. Abdom Imaging 2010; 35: Philip CA, Bisch C, Coulon A, de Saint-Hilaire P, Rudigoz RC, Dubernard G. Correlation between three-dimensional rectosonography and magnetic resonance imaging in the diagnosis of rectosigmoid endometriosis: a preliminary study on the first fifty cases. Eur J Obstet Gynecol Reprod Biol 2015; 187: Guerriero S, Alcazar JL, Ajossa S, Melis GB. Modified ultrasound scanning is a cost-effective method for the detection of deep infiltrating endometriosis. Fertil Steril 2009; 91: e38; author reply e Tammaa A, Fritzer N, Strunk G, Krell A, Salzer H, Hudelist G. Learning curve for the detection of pouch of Douglas obliteration and deep infiltrating endometriosis of the rectum. Hum Reprod 2014; 29: Scardapane A, Bettocchi S, Lorusso F, Stabile Ianora AA, Vimercati A, Ceci O, Lasciarrea M, Angelelli G. Diagnosis of colorectal endometriosis: contribution of contrast enhanced MR-colonography. Eur Radiol 2011; 21: Ultrasound Obstet Gynecol 2017; 49: