Does transvaginal ultrasonography combined with watercontrast in the rectum aid in the diagnosis of rectovaginal endometriosis infiltrating the bowel?

Transcription

1 Human Reproduction Vol.23, No.5 pp , 2008 Advance Access publication on February 29, 2008 doi: /humrep/den057 Does transvaginal ultrasonography combined with watercontrast in the rectum aid in the diagnosis of rectovaginal endometriosis infiltrating the bowel? M. Valenzano Menada, V. Remorgida, L.H. Abbamonte, A. Nicoletti, N. Ragni and S. Ferrero 1 Department of Obstetrics and Gynaecology, San Martino Hospital, University of Genoa, Largo R. Benzi 1, Genoa, Italy 1 Correspondence address. Tel: þ /þ ; Fax: þ ; dr@simoneferrero.com BACKGROUND: The aim of this study was to determine whether adding water-contrast in the rectum during transvaginal ultrasonography (RWC-TVS) improves the diagnosis of rectal infiltration in women with rectovaginal endometriosis. METHODS: This prospective study included 90 women, with suspect rectovaginal endometriosis, who underwent operative laparoscopy. TVS and RWC-TVS were independently performed by different investigators. RWC-TVS was performed by injecting saline solution into the rectal lumen under ultrasonographic control through a 6-mm catheter. Presence of rectovaginal nodules, presence and degree of rectal infiltration, and the largest diameter of the bowel nodules were evaluated. Ultrasonographic results were compared to surgical and histological findings. RESULTS: Although RWC-TVS had higher accuracy than TVS in diagnosing rectovaginal endometriosis, the difference between the two techniques was not statistically significant. RWC-TVS was significantly more accurate than TVS in determining the presence of endometriotic infiltration reaching at least the muscular layer of the rectal wall. The sensitivity of RWC-TVS in identifying rectal lesions was 97%, the specificity 100%, the positive predictive value 100% and the negative predictive value 91.3%. RWC-TVS caused a higher intensity of pain than TVS. CONCLUSIONS: RWC-TVS determines the presence of rectovaginal nodules infiltrating the rectal muscularis propria more accurately than TVS; RWC-TVS could be used when TVS cannot exclude the presence of rectal infiltration. Keywords: diagnosis; laparoscopy; rectal endometriosis; rectovaginal endometriosis; transvaginal ultrasonography Introduction Rectovaginal endometriosis involves the connective tissue between the anterior rectal wall and the vagina and it often infiltrates both. When endometriosis infiltrates the rectum, it may cause not only pain but also gastrointestinal symptoms including dyschezia, hematochezia, diarrhoea and constipation (Ferrero et al., 2005; Remorgida et al., 2005a). Surgical excision of rectovaginal endometriosis has been demonstrated to improve both pain symptoms and quality of life; however, the success rate depends on the complete excision of endometriosis, even when it infiltrates the bowel (Reich et al., 1991; Donnez et al., 1997; Redwine and Wright, 2001). Rectovaginal endometriosis is difficult to assess by clinical examination and infiltration of the rectal wall can only be suspected in 40 68% of the cases (Chapron et al., 2002; Bazot et al., 2003). Even during laparoscopy, generalist gynaecologists may fail to diagnose rectovaginal endometriosis (Griffiths et al., 2007). Therefore, imaging techniques are mandatory during the preoperative work-up. Determining before surgery whether bowel muscolaris is infiltrated by endometriosis, allows the gynaecologist to discuss the surgical approach (nodulectomy or bowel resection) with the colorectal surgeon. Furthermore, determining the presence and extension of rectal nodules allows the gynaecologist to obtain informed consent from the patient; this consent is particularly relevant when rectal resection is required, because the risk of complications increases (Remorgida et al., 2007). Several imaging methods have been used in the attempt to improve the non-invasive diagnosis of rectal infiltration in women with rectovaginal endometriosis. Colonoscopy is inaccurate in the diagnosis of bowel endometriosis because endometriotic lesions predominantly affect the serosa, muscularis and submucosa, while the mucosa is rarely involved (Zwas and Lyon, 1991; Redwine and Sharpe, 1995; Garry, 2004). Magnetic resonance imaging (MRI) has been reported to have a sensibility of 76.5% and a specificity of 97.9% for the diagnosis of rectal involvement in women with deep infiltrating endometriosis (Chapron et al., 2004). The use of the endorectal coil optimizes the finding of MRI (Kinkel et al., 1999), but its movements may be limited because of the pain induced by the pressure on the # The Author Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org 1069

2 Valenzano Menada et al. endometriotic nodules. In addition, the use of MRI in the diagnosis of endometriosis is based on the presence of haemorrhagic content within the lesion and rectal lesions often contain extensive fibrosis (Remorgida et al., 2005b) which may alter the signal pattern. Several studies proved the usefulness of rectal endoscopic ultrasonography for evaluating the infiltration of the intestinal wall by endometriosis (Chapron et al., 1998; Bazot et al., 2003); however, the quality of the examination is variable depending on the operator; furthermore, the gynaecologists who typically diagnose the disease often have limited experience in performing transrectal ultrasonography. Multislice computerized tomography combined with retrograde distension of the colon by rectal enteroclysis (MSCTe) has been proved to be effective in the diagnosis of bowel endometriosis (Biscaldi et al., 2007a, b). This technique reliably identifies endometriotic nodules located on the sigmoid colon, caecum and ileum; however, up to 11% of rectal nodules may be missed by MSCTe (Biscaldi et al., 2007a). Transvaginal ultrasonography (TVS) is the first line procedure for the exploration of the pelvic cavity. Bazot et al. (2003, 2007) showed that TVS reliably determines the presence and diameter of colorectal endometriotic lesions. On the basis of our previous experience of MSCTe, we recently applied the criteria of retrograde bowel distension to TVS. In a pilot study including 35 women, we showed that TVS combined with water-contrast in the rectum (RWC-TVS) is accurate in diagnosing rectal wall infiltration in women with rectovaginal endometriosis (Valenzano Menada et al., in press); the sensitivity of this technique for detecting rectal lesions reaching the muscular layer was 100%, the specificity 85.7%, the positive predictive value (PPV) 91.3% and the negative predictive value (NPV) 100%. The objective of the current study is to compare the effectiveness of TVS and RWC-TVS in the diagnosis of rectal infiltration in women with rectovaginal endometriosis. Materials and Methods This prospective study was performed in the period between October 2006 and November The study included women with suspect of rectovaginal endometriosis on the basis of pain symptoms (i.e. dysmenorrhoea, dyspareunia, chronic pelvic pain) and/or gynaecological examination. All women underwent operative laparoscopy at our hospital. Exclusion criteria comprised: patients who were virgins or who had any type of genital malformation that made physical examination or TVS impossible; previous surgical excision of bowel endometriosis. The study was approved by the local Institutional Review Board. Before surgery, all patients were informed about experimental use of RWC-TVS in the diagnosis of bowel endometriosis and signed a written informed consent form. Subjects of the study underwent laparoscopy and consented to bowel surgery on the basis of radiological investigations, in most cases MSCTe, which currently represents our standard for the diagnosis of intestinal endometriosis (Biscaldi et al., 2007a,b). TVS and RWC-TVS technique On the day before surgery, each patient was asked to drink four doses of a granular powder (Selg 1000 w ; Promefarm, Milan, Italy; each dose containing 5832 g polyethilenglycol 4000, 569 g anhydrous sodium sulphate, 1.69 g sodium bicarbonate, 1.46 g sodium chloride and 0.74 g potassium chloride) dissolved in 1000 ml of water per dose. A few hours before surgery, subjects of the study underwent TVS and RWC-TVS. The examinations were performed by using a Siemens Sonoline Antares ultrasound machine (Siemens Medical, Erlangen, Germany) with a MHz multi-frequency transvaginal probe. Two different experienced ultrasonographers independently performed the examinations. One operator performed all TVS; after the completion of this exam a second operator performed RWC-TVS. The operators were informed that rectovaginal endometriosis was suspected but they were unaware of the findings of vaginal or rectal examination and they were not informed of the findings of previous radiological examinations. The second operator was not aware of the findings of TVS at the time of RWC-TVS. The first operator (A.L.H) evaluated the presence of the rectovaginal nodule and its relations with the adjacent rectal wall. The examination particularly focused on the anterior and lateral sides of the rectum, where deep endometriotic nodules are usually located. The second operator (V.M.M.) immediately performed RWC-TVS and did not use TVS to verify the presence and characteristics of rectovaginal endometriosis. An assistant (N.A.) inserted a 6-mm (18 Ch) catheter into the rectal lumen up to a 20 cm distance from the anus. Saline solution was then injected inside the rectum under ultrasonographic control (Fig. 1). The exam was performed following distension of rectal wall. During the examinations the operators were asked to determine: the presence of the rectovaginal endometriotic nodule, the presence of infiltration of the rectal wall and the larger diameter of the bowel nodule. The findings of the exams were recorded by using a standardized form. Rectovaginal endometriosis appears ultrasonographically as rounded or triangular hypoechoic masses, located anterior or lateral to the rectum, immediately adjacent or close to the rectal wall. Rectal endometriotic infiltration was defined by the fact that the rectovaginal hypoechoic mass was adherent and/or penetrated into the intestinal wall thickening the muscolaris mucosa; hypoechoic or hyperechoic foci were sometimes present (Fig. 2). The longitudinal, anteroposterior and transversal axes of the lesions were measured and the larger diameter of the lesion was recorded. The examiners tried to identify the various layers of the rectal wall in order to estimate the depth of infiltration of the endometriotic lesion. Immediately after each exam, patients were asked to rate the discomfort encountered during TVS and RWC-TVS by using a 10-cm visual analogue scale (VAS); the left extreme represented the Figure 1: The three images (from left to right) show the progressive distension of the rectal wall during RWC-TVS 1070

3 RWC-TVS in rectal endometriosis Figure 2: On the left, TVS demonstrates a rectovaginal nodule (indicated by the arrow) which seems to infiltrate the muscularis of the bowel wall. On the right, RWC-TVS clearly demonstrates that the rectovaginal nodule infiltrates the rectal muscularis; the submucosa (hyperechoic, indicated by the asterisk) is infiltrated absence of pain, and the right extreme represented the worst possible pain. Mild pain was defined as VAS score,2, moderate pain as VAS score 2 and 5, and severe pain as VAS score.5. Surgical technique All surgical procedures were performed by a team of gynaecological and colorectal surgeons with extensive experience in the treatment of pelvic and bowel endometriosis (Remorgida et al., 2005a,b). The procedures were usually started laparoscopically; however, an initial vaginal approach was performed in some cases (Possover et al., 2000). At the beginning of surgery, an intrauterine manipulator was positioned in order to antevert the uterus. During the dissection of the posterior fornix of the vagina, the cleavage plane was identified by placing a ring forceps in the vagina and a probe in the rectum (Garry, 2004). The goal of the operation was to remove the disease en bloc; no attempt was made to dissect the endometriotic nodule from the rectum. After adequate adhesiolysis, the sigmoid colon and the rectum were systematically examined to verify the presence of endometriotic lesions. All visible bowel endometriotic lesions were removed. Bowel resection was performed according to previously published criteria (Remorgida et al., 2005a,b): single lesion.3 cm in diameter, single lesion infiltrating 50% of the bowel wall, three or more lesions infiltrating the muscular layer; in all the other cases, nodulectomy was performed. All nodulectomies (partial or full thickness) were performed at laparoscopy cutting the serosa around the tip of the nodule leaving at least 1 cm of macroscopically normal tissue; after the first incision, the nodule was removed following the cleavage plane. All bowel resections were started laparoscopically and the bowel was mobilized. Bowel resection was performed either by a laparotomic Pfannenstiel incision or by exteriorizing the bowel through a small suprapubic incision (3 5 cm). Segmental resection was performed with an automatic stapler. A Knight Griffen technique was used for low rectal lesions (Knight and Griffen, 1980; Griffen and Knight, 1984). Histological evaluation of bowel specimens Histopathologic criteria for the diagnosis of colorectal endometriosis were the presence of ectopic endometrial and stromal tissues penetrating through the bowel wall (Clement, 2002). The diameter of the endometriotic nodules was measured and recorded. The depth of infiltration of the endometriotic nodules was evaluated as previously described (Remorgida et al., 2005a; Biscaldi et al., 2007a). Statistical analysis Sensitivity, specificity, PPV, NPV and accuracy were determined for both TVS and RWC-TVS. The diagnostic value of each test also was evaluated using positive likelihood ratio (LRþ) and negative LR (LR2) (Deeks and Altman, 2004). Data distribution was evaluated by using the Shapiro Wilk test. Comparisons between multiple groups were performed by using the ANOVA on ranks, the post hoc Dunn s test was used for multiple comparisons. The McNemar s test with the Yates continuity correction was used to compare the accuracy of TVS and RWC-TVS in the diagnosis of rectovaginal endometriosis and rectal infiltration. The Pearson s correlation coefficient test was used to evaluate the correlation between the diameter of the endometriotic nodules estimated at TVS and RWC-TVS with that measured at the anatomopathological evaluation. Data were analysed using the Statistical Packages for the Social Sciences version 13.0 (SPSS, Chicago, IL, USA) and the Sigma-Stat 3.5 software package (SPSS). Results Demographic characteristics of the patients included in the study are listed in Table I. Surgery revealed that 69 patients (76.7%) had rectovaginal endometriotic nodules, 12 women (13.3%) had endometriotic lesions other than rectovaginal nodules and nine women (10.0%) did not have endometriosis. Rectal infiltration was observed in 29 women (32.2%). Bowel endometriotic nodules were excised by partial thickness nodulectomy in 5 cases, by full thickness nodulectomy in 5 cases and by segmental resection in 19 cases. Excision of pelvic endometriosis was performed by laparoscopy in all patients; all nodulectomies (either partial or full thickness) were performed by laparoscopy. Segmental bowel resections were performed by laparoscopy in 17 women and by laparotomy in 2 patients. The histological examination of specimens excised at surgery showed that rectal infiltration reached the serosa in 6 cases, the muscularis in 18 cases and the mucosa in 5 cases. Findings at TVS TVS correctly identified the presence of rectovaginal endometriosis in 64 (of 69) patients. In five women rectovaginal endometriotic nodules were not diagnosed at TVS, the largest diameter of the nodules excised at surgery in these patients was between 0.5 and 0.9 mm. In two women, a rectovaginal endometriotic nodule was diagnosed at TVS but not detected at surgery. The sensibility, specificity, PPV, NPV and accuracy of TVS in detecting the presence of rectovaginal endometriosis are given in Table II. TVS correctly identified the presence of rectal infiltration reaching at least the muscular layer in only 56.5% (13/23) of the cases. Table II shows the sensibility, specificity, PPV, NPV and accuracy of TVS in diagnosing endometriotic lesions reaching at least the muscular layer of the rectum. 1071

4 Valenzano Menada et al. Table I. Characteristics of the patients (n ¼ 90) included in the study. Age (years; median, range) 32 (18 42) Patients who used medical treatments for endometriosis (n, %) 82 (91.1) Oral contraceptive pill 55 (61.1) Vaginal ring 27 (30.0) Norethisterone acetate 23 (25.6) GnRH analogues 22 (24.4) Danazol 9 (10.0) Letrozole 4 (4.4) Endometriosis-related symptoms (n, %) Dysmenorrhoea 84 (93.3) Dyspareunia a 68 (82.9) Chronic pelvic pain 62 (68.9) Infertility 32 (39.0) Diarrhoea and/or constipation 61 (67.8) Bowel movement pain or cramping 69 (76.7) Pain on defecation 32 (35.6) Rectorrhagia 16 (17.8) Lower back pain 57 (63.3) a Only 82 women were sexually active. Findings at RWC-TVS RWC-TVS correctly identified the presence of rectovaginal endometriosis in 67 (of 69) women and the absence of rectovaginal nodules in 21 women. In two cases, small rectovaginal nodules (largest diameters 0.5 and 0.7 mm) were not diagnosed at RWC-TVS. The sensibility, specificity, PPV, NPV and accuracy of RWC-TVS in diagnosing rectovaginal endometriotic nodules are given in Table II. RWC-TVS had high sensitivity and specificity in identifying rectal infiltration reaching at least the muscular layer (Table II); however, it could not determine whether the mucosa was infiltrated. In patients with superficial lesions which did not reach the bowel muscularis, RWC-TVS did not reliably determine whether the bowel serosa was infiltrated (Table III). Comparison of TVS and RWC-TVS TVS and RWC-TVS correctly diagnosed the presence of rectovaginal endometriotic nodules in 83/90 (92.2%) and 88/90 (97.8%) patients, respectively. Although there was a trend for RWC-TVS to have higher accuracy than TVS in diagnosing rectovaginal endometriosis, the difference between the two techniques was not statistically significant (P ¼ 0.074). TVS and RWC-TVS correctly diagnosed the presence of endometriotic infiltration reaching at least the muscular layer of the rectal wall in 83.3 (75/90) and 98.9% (89/90) of the patients, respectively. The McNemar s test showed that the accuracy of RWC-TVS and TVS were significantly different in determining the presence of endometriotic nodules infiltrating the rectal wall (P ¼ 0.001). An example of such a diagnosis by RWC-TVS is shown in Fig. 3. Combining the results of the two techniques did not increase the sensitivity of RWC-TVS because the two false negative cases with RWC-TVS were not diagnosed by TVS. At histology the largest diameter of the endometriotic nodule ranged from 5 to 57 mm (mean + SD, mm), compared with 5 to 41 mm ( ) at TVS and 5 to 53 mm ( ) at RWC-TVS. Both TVS and RWC-TVS agreed with histology on the size of the lesions (P ¼ 0.01, r ¼ 0.912; P ¼ 0.01, r ¼ 0.986, respectively). Tolerability of RWC-TVS RWC-TVS determined higher intensity of pain on the VAS scale than TVS ( and , respectively; P, 0.001). However, it was possible to perform RWC-TVS in all the patients, and only 23 women (25.6%) defined the intensity of pain as severe (VAS score.5). For patients reporting higher intensity of pain at RWC-TVS, the median difference of pain intensity between the two exams was 1.8 (range, ). TVS determined a similar intensity of pain in women without endometriosis, in those with rectovaginal endometriosis without rectal infiltration and in women with rectovaginal nodules infiltrating the rectum (P ¼ 0.181). RWC-TVS determined significantly higher intensity of pain in women with rectovaginal nodules infiltrating the rectum than in those with only rectovaginal endometriosis (P, 0.05) and in those without endometriosis (P, 0.05). Women with rectovaginal nodules without rectal infiltration had higher intensity of pain at RWC-TVS than those without endometriosis (P, 0.05; Fig. 4). Discussion This study compared the accuracy of TVS and RWC-TVS in determining the presence of rectovaginal nodules and rectal infiltration in women with suspect of rectovaginal endometriosis. Our results indicate that adding water-contrast to the rectum improves the identification of rectal infiltration reaching at least the muscular layer. RWC-TVS identified 22 out of 23 cases of rectal infiltration (sensitivity, 95.7%); infiltration of the rectal muscularis is particularly relevant because it may increase the risks of post-operative complications if the endometriotic nodule is completely excised. Table II. Accuracy of TVS and RWC-TVS in detecting the presence of rectovaginal endometriosis. Sensitivity, % Specificity, % PPV, % NPV, % Accuracy, % LRþ LR2 Diagnosis of rectovaginal endometriosis TVS RWC-TVS a 0.03 Diagnosis of rectal infiltration TVS RWC-TVS a 0.04 a LRþ could not be calculated because of the absence of false positive cases. 1072

5 RWC-TVS in rectal endometriosis Table III. Comparison between depth of infiltration at histology with findings at TVS and RWC-TVS. Histology TVS RWC-TVS Depth of infiltration n Infiltration of the serosa 6 Absence of rectal infiltration, n ¼ 4 Absence of rectal infiltration, n ¼ 4 Infiltration of the serosa, n ¼ 1 Infiltration of the serosa, n ¼ 2 Infiltration of the muscularis, n ¼ 1 Infiltration of the muscularis 18 Absence of rectal infiltration, n ¼ 5 Infiltration of the serosa, n ¼ 1 Infiltration of the serosa, n ¼ 3 Infiltration of the muscularis, n ¼ 15 Infiltration of the muscularis, n ¼ 8 Infiltration of the mucosa, n ¼ 2 Infiltration of the mucosa, n ¼ 2 Infiltration of the mucosa 5 Absence of rectal infiltration, n ¼ 2 Infiltration of the muscularis, n ¼ 3 Infiltration of the muscularis, n ¼ 3 Infiltration of the mucosa, n ¼ 2 Figure 3: A large rectovaginal nodule infiltrating the bowel muscularis (indicated by the asterisk) is demonstrated by RWC-TVS (on the left). Surgery confirmed the presence of the nodule (central panel) which was removed by laparoscopic bowel resection. Histology confirmed the endometriotic infiltration of the rectal muscularis (on the right) The use of transvaginal ultrasound for diagnosing rectal infiltration in women with rectovaginal endometriosis was originally proposed by Gorell et al. (1989). More recently, in a series of six women with rectovaginal endometriosis, Koga et al. (2003) demonstrated that both transvaginal and transrectal ultrasonography provides characteristic ultrasonographic images of rectosigmoid nodules; however, the authors did not determine the sensitivity and sensibility of these techniques in estimating bowel infiltration. By using TVS, Bazot et al. (2003) diagnosed endometriotic infiltration of the rectal muscularis propria in 21 out of 22 cases; the exam had a sensitivity of 95.5%, a specificity of 100%, a PPV of 100% and a NPV of 88.9%. By combining TVS with a retrograde bowel preparation performed 1 h before the examination, Abrão et al. (2005) obtained a sensitivity of 95.4%, a specificity of 96.4%, a PPV of 95.4% and a NPV of 96.5% in diagnosing rectal infiltration in 110 women with deep infiltrating endometriosis. In a series of 46 patients, sonovaginography, a technique combining TVS with the introduction of a saline solution in the vagina, was demonstrated to diagnose rectovaginal endometriosis more accurately than TVS alone; unfortunately, only three patients included in the study had infiltration of the rectal wall and therefore the value of this technique in the diagnosis of rectal infiltration remains unclear (Dessole et al., 2003). More recently, a tenderness-guided TVS has been proposed for the detection of deep endometriosis. This technique, which consists of the introduction of 12 ml of ultrasound transmission gel in the probe cover, was proved to be accurate in the diagnosis of deep endometriosis (Guerriero et al., 2007). Unfortunately, only four patients with infiltration of the rectal wall were included in the study and therefore the value of tenderness-guided TVS in determining bowel involvement remains to be determined. Our study was performed in a Figure 4: Intensity of pain measured on a 10 cm VAS scale during TVS and RWC-TVS in women without endometriosis (1), women with rectovaginal endometriosis without rectal infiltration (2) and in women with rectovaginal nodules infiltrating the rectum (3) referral centre for the surgical treatment of endometriosis. In this setting, the removal of rectovaginal endometriosis is considered a routinary procedure and the primary aim of RWC-TVS was to exclude the presence of rectal infiltration of the muscular layer which increases the risk post-operative complications and requires adequate preoperative consent (Remorgida et al., 2007). The very low LR- of RWC-TVS in identifying endometriotic nodules infiltrating the rectal muscularis demonstrates the effectiveness of this technique in excluding the presence of rectal infiltration reaching at least the muscular layer; on the contrary, TVS determined only a minimal decrease in the likelihood of having rectal infiltration (Table II). The use of RWC-TVS in diagnosing rectal infiltration in women with rectovaginal endometriosis has several potential advantages over other techniques. First, TVS is a simple technique which has high diffusion and it is usually performed by gynaecologists; the longitudinal images obtained at RWC-TVS 1073

6 Valenzano Menada et al. are more familiar to the gynaecologist than those obtained by rectal endoscopic ultrasonography. Secondly, this examination can be easily performed in any gynaecological department, while the equipment for transrectal endoscopic ultrasound in often unavailable (Abrão et al., 2007). Thirdly, RWC-TVS is less invasive than other examinations (such as rectal endoscopic ultrasonography) which sometimes necessitate general anesthesia (Delpy et al., 2005). Although the retrograde distension of the rectal lumen may cause variable intensity of pain in different patients (Fig. 4), it was possible to complete the exam in all women included in the current study without any anesthesia. Fourthly, RWC-TVS may provide information on the distensibility of the intestinal walls. Unfortunately, bowel distensibility was not systematically evaluated and recorded in the current study; however, this information could be useful for the surgeon because distensibility may reflect endometriosis-associated fibrosis present in the intestinal wall (Remorgida et al., 2005b). Finally, RWC-TVS is cheaper than other investigations such as MRI and MSCTe. A limitation of RWC-TVS consists of the fact that it is unsuitable for diagnosing endometriotic nodules located above the rectosigmoid junction which are beyond the field of view of ultrasonography; this consideration is particularly relevant in the light of the multifocal nature of bowel endometriosis (Kavallaris et al., 2003). Therefore, even if future studies will confirm our findings, it is unlikely that RWC-TVS may completely replace more expensive imaging techniques (such as MRI or MSCTe). On the contrary, combining RWC-TVS with MSCTe may improve the preoperative detection of all bowel endometriotic lesions. An obvious limitation of RWC-TVS consists of the fact that it is an operator-dependent exam; an experienced physician (with over 10 years experience in TVS) performed RWC-TVS in this study. It remains unclear whether this exam may have the same accuracy in the hands of less experienced physicians. We are aware that some limitations characterize the current study. RWC-TVS was performed on the day of surgery, after the patients received mechanical bowel preparation which removed bulky intraluminal contents and may have improved the visualization of the bowel wall. This preparation is unsuitable for a routine ultrasound examination; however, using a simple rectal enema a few hours before the exam may allow obtaining similar results. In addition, the anxiety of the patients on the day of surgery may have biased the evaluation of pain intensity during the exams. In the present study, findings of RWC-TVS and TVS were compared with surgery and histology. Obviously, an incomplete surgical excision of bowel nodules at the time of partial thickness nodulectomy may determine a bias in the findings. However, the surgeons had extensive experience in the treatment of bowel endometriosis and their aim was to completely excise bowel endometriosis. In addition, the majority of patients included in the study underwent either bowel resection or full thickness disk resection thus allowing the pathologist to reliably estimate the depth of infiltration of the endometriotic lesions in the bowel wall. Another criticism of our study is the fact that it was performed in a referral centre for the surgical treatment of endometriosis and we evaluated the efficacy of RWC-TVS in a population of 1074 highly symptomatic women (Table I) which had high risk of rectal infiltration. The sensibility and specificity of this technique should be confirmed by further investigations in populations with lower prevalence of bowel endometriosis. This study was specifically designed to determine the effectiveness of RWC-TVS in diagnosing rectovaginal endometriosis and, in particular, rectal infiltration; therefore the presence of other endometriotic lesions was not systematically investigated and recorded during this study. This choice is justified by the fact that the excision of vaginal, vesical and periureteral endometriotic lesions is considered a routinary procedure at our centre and it is associated with a very low post-operative complication rate. On the contrary, the excision of rectovaginal nodules infiltrating the bowel may require the assistance of a general surgeon and it is associated with a higher risk of complications (Remorgida et al., 2005a, 2007). In addition, we doubt that RWC-TVS may increase the effectiveness of TVS in diagnosing endometriotic lesions infiltrating the uterosacral ligaments, the vagina and the bladder. In two patients, rectovaginal endometriotic nodules were diagnosed at TVS, but they were not detected at RWC-TVS and at surgery. It seems unlikely that the surgeon failed to diagnose these nodules during surgery; in fact, the rectovaginal septum was dissected and normal areolar tissue of the rectovaginal septum was clearly identified. In addition, the outline of the posterior vaginal vault was clearly defined by placing ring forceps in the posterior vaginal fornix and pushing them towards the patient s head (Garry, 2004). In summary, this study shows that RWC-TVS is more accurate than TVS in diagnosing rectal infiltration reaching at least the muscularis propria in women with rectovaginal endometriosis. However, this exam cannot determine whether the infiltration reaches the rectal submucosa. RWC-TVS may be more painful than TVS, therefore it could be used when TVS cannot exclude the presence of rectal infiltration in women with rectovaginal endometriosis. Future studies should compare RWC-TVS with other techniques currently used in the diagnosis of bowel endometriosis, in particular, rectal endoscopic ultrasonography. Declaration of authors roles M.V.M., L.H.A. and A.N. performed the ultrasonographic examinations; V.R. and S.F. performed the surgical procedures. L.H.A. and A.N. prospectively collected the clinical data in an Excel file. M.V.M., V.R., S.F. and N.R. designed the study. The original draft of the manuscript was written by S.F. and revised by M.V.M., V.R. and N.R. References Abrão MS, Gonçalves MO, Gonzales M, Dias JA Jr, Podgaec S. It is possible to evaluate deeply infiltrating endometriosis with transvaginal ultrasound. Eur J Obstet Gynecol Reprod Biol 2005;123(Suppl 1):S14. Abrão MS, Gonçalves MO, Dias JA, Jr, Podgaec S, Chamie LP, Blasbalg R. Comparison between clinical examination, transvaginal sonography and magnetic resonance imaging for the diagnosis of deep endometriosis. Hum Reprod 2007;22:

7 RWC-TVS in rectal endometriosis Bazot M, Detchev R, Cortez A, Amouyal P, Uzan S, Darai E. Transvaginal sonography and rectal endoscopic sonography for the assessment of pelvic endometriosis: a preliminary comparison. Hum Reprod 2003;18: Bazot M, Malzy P, Cortez A, Roseau G, Amouyal P, Daraï E. Accuracy of transvaginal sonography and rectal endoscopic sonography in the diagnosis of deep infiltrating endometriosis. Ultrasound Obstet Gynecol 2007;30: Biscaldi E, Ferrero S, Fulcheri E, Ragni N, Remorgida V, Rollandi GA. Multislice CT enteroclysis in the diagnosis of bowel endometriosis. Eur Radiol 2007a;17: Biscaldi E, Ferrero S, Remorgida V, Rollandi GA. Bowel endometriosis: CT-enteroclysis. Abdom Imaging 2007b;32: Chapron C, Dumontier I, Dousset B, Fritel X, Tardif D, Roseau G, Chaussade S, Couturier D, Dubuisson JB. Results and role of rectal endoscopic ultrasonography for patients with deep pelvic endometriosis. Hum Reprod 1998;13: Chapron C, Dubuisson JB, Pansini V, Vieira M, Fauconnier A, Barakat H, Dousset B. Routine clinical examination is not sufficient for diagnosing and locating deeply infiltrating endometriosis. J Am Assoc Gynecol Laparosc 2002;9: Chapron C, Vieira M, Chopin N, Balleyguier C, Barakat H, Dumontier I, Roseau G, Fauconnier A, Foulot H, Dousset B. Accuracy of rectal endoscopic ultrasonography and magnetic resonance imaging in the diagnosis of rectal involvement for patients presenting with deeply infiltrating endometriosis. Ultrasound Obstet Gynecol 2004;24: Clement MD. Diseases of the peritoneum (including endometriosis). In: Kurman RJ (ed). Blaustein s Pathology of the Female Genital Tract. New York, USA: Springer-Verlag, 2002, Deeks JJ, Altman DG. Diagnostic tests 4: likelihood ratios. BMJ 2004;329: Delpy R, Barthet M, Gasmi M, Berdah S, Shojai R, Desjeux A, Boubli L, Grimaud JC. Value of endorectal ultrasonography for diagnosing rectovaginal septal endometriosis infiltrating the rectum. Endoscopy 2005;37: Dessole S, Farina M, Rubattu G, Cosmi E, Ambrosini G, Nardelli GB. Sonovaginography is a new technique for assessing rectovaginal endometriosis. Fertil Steril 2003;79: Donnez J, Nisolle M, Gillerot S, Smets M, Bassil S, Casanas-Roux F. Rectovaginal septum adenomyotic nodules: a series of 500 cases. BJOG 1997;104: Ferrero S, Abbamonte LH, Remorgida V, Ragni N. Abdominal pain, bloating, and urgency. Obstet Gynecol 2005;106:195. Garry R. Electrosurgical resection of endometriosis. In: Redwine DB (ed). Surgical Management of Endometriosis. Martin Dunitz. London, UK: Taylor and Francis Group, 2004, Gorell HA, Cyr DR, Wang KY, Greer BE. Rectosigmoid endometriosis. Diagnosis using endovaginal sonography. JUltrasoundMed1989;8: Griffen FD, Knight CD. Stapling technique for primary and secondary rectal anastomoses. Surg Clin North Am 1984;64: Griffiths AN, Koutsouridou RN, Penketh RJ. Rectovaginal endometriosis - a frequently missed diagnosis. J Obstet Gynaecol 2007;27: 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: Kavallaris A, Kohler C, Kuhne-Heid R, Schneider A. Histopathological extent of rectal invasion by rectovaginal endometriosis. Hum Reprod 2003;18: Kinkel K, Chapron C, Balleyguier C, Fritel X, Dubuisson JB, Moreau JF. Magnetic resonance imaging characteristics of deep endometriosis. Hum Reprod 1999;14: Knight CD, Griffen FD. An improved technique for low anterior resection of the rectum using the EEA stapler. Surgery 1980;88: Koga K, Osuga Y, Yano T, Momoeda M, Yoshino O, Hirota Y, Kugu K, Nishii O, Tsutsumi O, Taketani Y. Characteristic images of deeply infiltrating rectosigmoid endometriosis on transvaginal and transrectal ultrasonography. Hum Reprod 2003;18: Possover M, Diebolder H, Plaul K, Schneider A. Laparascopically assisted vaginal resection of rectovaginal endometriosis. Obstet Gynecol 2000;96: Redwine DB, Sharpe DR. Laparoscopic surgery for intestinal and urinary endometriosis. Baillieres Clin Obstet Gynaecol 1995;9: Redwine DB, Wright JT. Laparoscopic treatment of complete obliteration of the cul-de-sac associated with endometriosis: long-term follow-up of en bloc resection. Fertil Steril 2001;76: Reich H, McGlynn F, Salvat J. Laparoscopic treatment of cul-de-sac obliteration secondary to retrocervical deep fibrotic endometriosis. J Reprod Med 1991;36: Remorgida V, Ragni N, Ferrero S, Anserini P, Torelli P, Fulcheri E. The involvement of the interstitial Cajal cells and the enteric nervous system in bowel endometriosis. Hum Reprod 2005a;20: Remorgida V, Ragni N, Ferrero S, Anserini P, Torelli P, Fulcheri E. How complete is full thickness disc resection of bowel endometriotic lesions? A prospective surgical and histological study. Hum Reprod 2005b;20: Remorgida V, Ferrero S, Fulcheri E, Ragni N, Martin DC. Bowel endometriosis: presentation, diagnosis, and treatment. Obstet Gynecol Surv 2007;62: Menada M, 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 (in press) (Epub ahead of print). Zwas FR, Lyon DT. Endometriosis. An important condition in clinical gastroenterology. Dig Dis Sci 1991;36: Submitted on December 25, 2007; resubmitted on January 30, 2008; accepted on February 7,