Ultrasound Obstet Gynecol 2012; 39: 336–340 Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/uog.10052

Endometriosis of the abdominal wall: ultrasonographic and Doppler characteristics L. SAVELLI*, L. MANUZZI*, N. DI DONATO*, N. SALFI†, G. TRIVELLA‡, M. CECCARONI‡ and R. SERACCHIOLI* *Gynecology and Reproductive Medicine Unit, S.Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy; †Department of Pathology, S.Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy; ‡Department of Obstetrics and Gynecology, Sacred Heart Hospital, Negrar, Italy

K E Y W O R D S: abdominal wall; accuracy; Doppler; endometriosis; transvaginal sonography; ultrasound

ABSTRACT Objectives To describe the sonographic and clinical features of abdominal wall endometriosis (AWE), a frequently misdiagnosed condition. Methods This was a retrospective study of 21 consecutive women with pathologically proven endometriosis of the abdominal wall. Ultrasonographic and Doppler examinations were performed, before surgery, with a high-frequency linear transducer. The clinical data and the results of the sonographic examinations were reviewed and described. Results At ultrasound, all the nodules appeared as discrete solid masses that were less echogenic than the surrounding hyperechoic fat. The nodules had a median diameter of 20 (range, 5–50) mm and in 18/21 (86%) cases the nodules had a round/oval shape. In eight of 21 (38%) women the AWE was located at the umbilicus, in six of 21 (29%) it was between the transverse suprapubic line and the umbilicus, in five of 21 (24%) it was found along the scar of a previous Cesarean section and in two of 21 (9%) it was in the right inguinal canal. The content was homogeneously hypoechoic in 12/21 (57%) women and inhomogeneous in the other nine (43%). The outer borders were invariably ill defined. Scarce blood vessels were found by power Doppler. Cyclic or continuous spontaneous pain at the level of the AWE was present in 19/21 (91%) cases, and two (9%) patients were asymptomatic. Conclusions Hypoechoic round/oval nodules with illdefined borders and a hyperechoic rim should raise the suspicion of abdominal wall endometriosis, even in patients with no history of endometriosis or previous laparotomic surgery. Pressing the ultrasound probe

against the nodule should reinforce a suspected diagnosis because of the pain it induces. Copyright  2012 ISUOG. Published by John Wiley & Sons, Ltd.

INTRODUCTION Endometriosis is defined as the presence of ectopic, functioning endometrial tissue outside the uterine cavity. In most cases it is located within the pelvis, but it can even be found in the lung, bowel, ureter and abdominal wall1 . The expression ‘abdominal wall endometriosis (AWE)’ is used to indicate the presence of ectopic endometrium located far from the peritoneum, embedded in the subcutaneous fatty tissue and the abdominal wall muscle layers. This entity is considered rare; nonetheless, given an estimated incidence of 0.03–1% after Cesarean section2 , its incidence is expected to rise in many countries. Many cases of AWE occur after laparoscopic or laparotomic surgery involving the uterine cavity; the majority of AWE reported have been described as being adjacent to Cesarean-section scars1,3,4 . However, this definition also includes lesions that are not a consequence of any previous surgery. The symptoms of AWE are non-specific and include cyclic or continuous pain associated with a palpable mass4 – 6 . Patients may complain of signs and symptoms suggestive of concomitant pelvic endometriosis. Abdominal wall endometriosis is often misdiagnosed as a hernia, suture granuloma, primary or metastatic tumor, hematoma or lipoma of the abdominal wall, thus resulting in unexpected findings at surgery1 . A correct preoperative diagnosis would help in counseling the patient and in planning appropriate surgery; knowledge of the nodule size and its extension through the abdominal muscular

Correspondence to: Dr L. Savelli, Via Pietro Mengoli, 31/4, 40138 Bologna, Italy (e-mail: [email protected]) Accepted: 4 July 2011

Copyright  2012 ISUOG. Published by John Wiley & Sons, Ltd.

ORIGINAL PAPER

Abdominal wall endometriosis fascia would help in choosing the best method for closing the fascia defect (e.g. the use of a mesh). The preoperative work-up of women with a palpable mass in the abdominal wall usually includes transabdominal ultrasound with sonographically guided fine-needle aspiration biopsy and cytologic analysis, and computed tomography (CT) or magnetic resonance imaging (MRI). Unfortunately, AWE has received little attention in the radiological literature, and the elevated number of misdiagnoses reflects the scant data available regarding its sonographic features. Transabdominal sonography using a linear transducer has been found to be useful in detecting and locating AWE, but the sonographic pattern of surgically confirmed AWE has still not been described in a large population of women. Previous literature on this subject is limited to case reports3,7 , the largest series of which are those reporting the sonographic features of AWE located on the scars of a previous Cesarean section7,8 . Unfortunately, no data are available on AWE that develops spontaneously in intact abdominal walls. Moreover, scant data are available on the results of preoperative ultrasound examinations using a linear high-frequency transducer. In this study, we retrospectively evaluated and described the preoperative sonographic findings in women with surgically proven AWE in order to assess the spectrum of presurgical findings.

METHODS We conducted a computerized search of the pathology databases in the two participating hospitals, both tertiary referral units for the diagnosis and treatment of endometriosis. In the period between January 2001 and March 2007, 30 consecutive women with pathologically proven endometriosis of the abdominal wall were retrospectively selected. Nine patients were excluded because of inconclusive histological reports, a lack of digital ultrasonographic images or when only a transabdominal convex probe was used. Twenty-one patients constituted the study population. Fifteen patients were operated on in Bologna, at the Department of Gynecology and Reproductive Medicine Unit (S. OrsolaMalpighi Hospital), and six patients were operated on in Negrar, at the Sacro Cuore Don Calabria Hospital. All patients had given their informed consent after reading precise information about the aims of the procedure. All scans were performed, within a 1-month period before surgery, by two investigators (L.S. in Bologna and G.T. in Negrar) who were aware of the patients’ clinical history and symptoms. Ultrasonographic examinations were performed using commercially available sonographic machines (Esaote Technos MP, Genoa, Italy; Voluson E8, GE Medical Systems, Zipf, Austria; and Toshiba Aplio, Toshiba Medical Systems, Nasu, Japan) equipped with a multifrequency 5.0–13.0-MHz linear transducer. Power Doppler sonography, with a pulse-repetition frequency of 500–750 Hz, was used to assess the vascularity of all lesions. For each

Copyright  2012 ISUOG. Published by John Wiley & Sons, Ltd.

337

lesion, the three orthogonal diameters were recorded and the mean was calculated; the shape, the location (e.g. along the scar of a previous Cesarean section, at the level of the umbilicus or at the inguinal canal), the depth (e.g. superficial in the subcutaneous tissue or involving the muscle layer), the echotexture (echogenicity was compared with that of adjacent normal subcutaneous tissue) and the appearance of the margins (smooth, irregular or frankly spiculated) and of the surrounding tissue were recorded. The physicians performing the examination gave a subjective impression of scarce vascularization when only one to three vascular spots (of arterial or venous origin) were found within the nodule, moderate when four to six vascular spots were found and abundant if more than seven vascular spots could be seen. A detailed transabdominal and transvaginal examination of the pelvis was always carried out with a convex 3.5–5.0-MHz transducer and an endocavitary 5.0–9.0-MHz transducer to evaluate the anatomy of the uterus and the ovaries. Digital images and videos of all the patients were subsequently reviewed by two authors (L.S. and G.T.) who agreed on their classification and description. All patients underwent surgery at the same institution where the sonography was carried out, with a complete resection of the mass and histological examination of the specimen. A diagnosis of AWE was made at histology when endometrial glands and stroma were found within the soft tissues and muscles of the abdominal wall, circumscribed by an inflammatory process and fibrosis. All clinical and ultrasound information was entered into a dedicated Microsoft Excel file (Microsoft Corp., Redmond, WA, USA) that was used for statistical analysis.

RESULTS The clinical characteristics of the 21 women (mean age 36 years; range, 25–48 years) constituting the study population are given in Table 1. Nine patients were excluded because of inconclusive histologic or ultrasonographic data, a lack of digital interpretable images or when only a transabdominal convex probe was used. In 19 (91%) cases, cyclic or continuous spontaneous abdominal pain was reported, and two (9%) patients were asymptomatic; only if pressure was exerted on the nodule was slight pain evoked in these patients. The associated symptoms suggestive of ovarian or pelvic endometriosis were: dysmenorrhea in 11 (52%) patients, dyspareunia in seven (33%) and infertility in six (29%). In eighteen (86%) patients, the nodule was palpable; in one of these patients, cutaneous bleeding was occasionally present. The indications for ultrasound examination were: pain (15/21 cases; 71%), a palpable mass (7/21 cases; 33%) and suspicion of umbilical hernia (4/21 cases; 19%); five patients had more than one indication. Overall, nine (43%) patients had previously undergone surgery (six by laparoscopy and three by laparotomy) for ovarian or pelvic deepinfiltrating endometriosis.

Ultrasound Obstet Gynecol 2012; 39: 336–340.

Savelli et al.

338 Table 1 Clinical characteristics of 21 patients who underwent surgery for abdominal wall endometriosis Clinical characteristic Cyclic or continuous pain Palpable mass Asymptomatic Pain at palpation Spontaneous cyclic bleeding Previous surgery for endometriosis Previous Cesarean section Symptoms suggestive for pelvic endometriosis Dysmenorrhea Dyspareunia Infertility

n (%)

19 (91) 18 (86) 2 (9) 21 (100) 1 (5) 9 (43) 8 (38) 11 (52) 7 (33) 6 (29)

Table 2 Sonographic features of abdominal wall endometriosis Sonographic feature Diameter (mm) Shape Round/oval Stellate/irregular Location Umbilicus Cesarean section scar Between transverse suprapubic line and umbilicus Inguinal canal Depth Superficial (subcutaneous tissue only) Deep (involving the muscle layer) Involving superficial and deep layers Echotexture Homogeneously hypoechoic Heterogeneous Margins Ill-defined, blurred Smooth Vascularization at color Doppler Scarce blood vessels Abundant blood vessels Pain at pressure with probe

Figure 1 Transverse sonogram of the abdominal wall showing a 20-mm abdominal wall endometriosis nodule with hypoechoic content and blurred outer margins (calipers). The nodule is enclosed in the subcutaneous fatty tissue above the muscular fascia, along the scar of a previous Cesarean section.

Value 20 (5–50) 18/21 (86) 3/21 (14) 8/21 (38) 5/21 (24) 6/21 (29) 2/21 (10) 12/21 (57) 4/21 (19) 5/21 (24) 12/21 (57) 9/21 (43) 21/21 (100) 0/21 21/21 (100) 0/21 21/21 (100)

Figure 2 Abdominal wall endometriosis of the right inguinal canal, appearing as a small hypoechoic nodule along the course of the inguinal canal.

Data are given as median (range) or n (%).

In all the women, ultrasound examination of the AWE showed the presence of a fixed solid hypoechoic mass (less echogenic than the surrounding hyperechoic fat) within the abdominal wall. Eighteen patients had a single nodule, one patient had two nodules (one large nodule associated with a small lesion) and two patients had multiple adjacent endometriotic nodules (in both cases a large nodule was found surrounded by small indefinite satellite lesions). Whenever more than one nodule was found in a patient, the larger one was taken as the most representative and was included in the analysis. The median diameter of the nodules was 20 (range, 5–50) mm. The sonographic characteristics of the nodules are given in Table 2. The AWE had a round/oval shape in 18/21 (86%) cases and a stellate morphology in three of 21 (14%) cases (Figure 1). In eight of 21 (38%) cases, the AWE was found at the level of the umbilicus (six of these women had

Copyright  2012 ISUOG. Published by John Wiley & Sons, Ltd.

previously undergone laparoscopy for ovarian or pelvic endometriosis); in five of 21 (24%), it was located along the scar of a Cesarean section. In six of 21 (29%) women, the nodule was found between the transverse suprapubic line and the umbilicus, whereas in two of 21 (10%) cases, the nodule was in the right inguinal canal (Figure 2) (in one of these patients the nodule was associated with an inguinal hernia). The nodule was located above the abdominal wall fascia muscularis in 12/21 (57%) cases and below it in four of 21 (19%), whereas in five of 21 (24%) cases the nodule extended grossly through the abdominal wall fascia muscularis. The echotexture was inhomogeneous in nine of 21 (43%) cases owing to the presence of inner hyperechoic punctate echoes or short bright strands (seven women; 33%), or small cystic lacunae (two women; 9%); in the remaining 12/21 (57%) cases, the AWE was homogeneously hypoechoic. In all 21 women, the AWE had

Ultrasound Obstet Gynecol 2012; 39: 336–340.

Abdominal wall endometriosis

339

Figure 3 Transverse sonogram of the abdominal wall showing a stellate hypoechoic nodule with spiculated margins located at the level of the umbilicus. Note the absence of blood vessels within the nodule at Doppler examination.

ill-defined, blurred outer borders surrounded by a hyperechoic ring (Videoclip S1). Power Doppler examination invariably revealed scarce blood vessels within the AWE (Figure 3). In terms of sonographic features, no differences could be found in small vs. large areas of AWE.

Figure 4 Histological specimen of abdominal wall endometriosis (Mallory staining, original magnification × 10). The nodule shows scarce endometrial glands and stroma (arrows) embedded within skeletal muscle cells (red) and collagen fibers (blue).

DISCUSSION Endometriosis arising in the abdominal wall represents a diagnostic challenge. Its low prevalence, together with the scant data available regarding its sonographic features, are possible causes of misdiagnosis. At ultrasound examination, AWE shows features similar to those of deep infiltrating pelvic endometriosis but different from those of ovarian endometriotic cysts. In fact, the nodules appear as solid (only two cases had some cystic areas within the AWE) with ill-defined, blurred outer borders. At histological examination, this appearance corresponds to the presence of scarce endometrial glands and stroma embedded within fibroblasts, collagen fibers and skeletal muscle cells (Figure 4). The hyperechoic ring surrounding the nodule corresponds to adipose tissue that has become edematous and filled with cells of inflammatory origin (histiocytes and granulocytes). Power Doppler disclosed only scarce blood vessels within the nodule in all patients; this appears to be in contrast to previous studies on AWE occurring in the region of a Cesarean section7,8 , which reported a high proportion of strongly vascularized lesions. In the vast majority of previously reported cases of AWE, the nodules were found at the site of a previous operation (laparotomic or laparoscopic), with a Cesareansection scar being the most frequently described location of the nodules. In our series, we reported several cases of spontaneous AWE that were not located on a previous abdominal incision, and this was in agreement with a large surgical/pathologic review of 445 cases1 in which 20% of cases of AWE were not associated with abdominal scars. In our population, out of eight cases of umbilical

Copyright  2012 ISUOG. Published by John Wiley & Sons, Ltd.

Figure 5 When a bulky abdominal wall endometriosis nodule is removed, a wide laparotomy may be required as well as a large resection of the abdominal fascia (arrows). A non-resorbable mesh may be required to close the defect.

endometriosis, two women had not previously been operated on (both had associated ovarian and pelvic endometriosis). Moreover, in six women, the nodule was found at various levels between the transverse suprapubic line and the umbilicus. Three of these patients had a history of one or more Cesarean sections, but the Cesarean section scar was free from endometriosis and the AWE was found in the subcutaneous tissue; therefore, it is plausible that the Cesarean sections were unrelated to the etiology of the lesions. Two women had AWE in the inguinal canal (one had a history of several laparoscopic operations for severe endometriosis). Therefore, a total of six of 21 (29%) women were diagnosed as having AWE in

Ultrasound Obstet Gynecol 2012; 39: 336–340.

Savelli et al.

340

a scarless abdomen. While the presence of endometriosis at the level of previous abdominal incisions can easily be explained by the displacement of endometrial cells or endometriotic islets, spontaneous AWE in a scarless abdomen must have another pathogenesis. We hypothesized that regurgitated endometrial cells can grow spontaneously at specific anatomic sites, such as the umbilicus and the inguinal canal, permitting their seeding and implantation. Moreover, individual endometrial cells can escape from the uterus through lymphovascular channels and enter the peripheral circulation, reaching ectopic sites1 . As far as the scanning technique is concerned, we recommend proper magnification of the image in order to reduce the field of view to 3–5 cm in depth. One should then distinguish the normal abdominal layers on a transverse or longitudinal section far from the site where the AWE is suspected. We suggest locating the skin layer, the underlying hyperechoic subcutaneous tissue and the hypoechoic muscle layer covered by the thin hyperechoic abdominal fascia. Underneath the muscle layer, a variable amount of subperitoneal fat can be present, delimited by the inferior parietal peritoneum, easily seen as a thin, hyperechoic line located above the intestine. Then, sliding the probe while exerting slight pressure will help to locate the AWE, as a result of the pain induced. Ultrasound examination should focus on estimating the size of the nodule and evaluating whether or not the AWE infiltrates the abdominal fascia. In fact, small lesions found in the subcutaneous tissue are relatively easy to remove. Those lesions which are adherent to the rectus sheath or extend into the muscle layer will probably require a wider laparotomy in order to be completely resected, and the risks and recovery time will accordingly be greater. Mesh repairs may be required to fill large defects in the fascia, and these procedures require adequate counseling and precise surgical planning (Figure 5). Knowledge of the specific sonographic features of AWE would help

to differentiate AWE from other abdominal-wall masses, such as hernias, hematomas, lymphomas, lipomas and subcutaneous cysts. Additional studies, such as fine-needle aspiration, CT and MRI should be carried out in the few cases in which diagnosis is still in doubt. The limitations of our study are its retrospective nature and the fact that no standardized protocol was predefined in order to scan the patients; this might have led to some lack of uniformity. Nonetheless, all lesions were evaluated using a linear high-frequency transducer by physicians working in well-recognized tertiary-care hospitals for the diagnosis and treatment of endometriosis. Moreover, a consensus was reached by the authors on every case, by reviewing several digital images and videos. This should have minimized the limitations and reinforced the overall data analysis.

REFERENCES 1. Horton JD, DeZee KJ, Ahnfeldt EP, Wagner M. Abdominal wall endometriosis: a surgeon’s perspective and review of 445 cases. Am J Surg 2008; 196: 207–212. 2. Dwivedi AJ, Agrawal SN, Silva YJ. Abdominal wall endometriomas. Dig Dis Sci 2002; 47: 456–461. 3. Hensen JH, Van Breda Vriesman AC, Puylaert JB. Abdominal wall endometriosis: clinical presentation and imaging features with emphasis on sonography. AJR Am J Roentgenol 2006; 186: 616–620. 4. Blanco RG, Parithivel VS, Shah AK, Gumbs MA, Schein M, Gerst PH. Abdominal wall endometriomas. Am J Surg 2003; 185: 596–598. 5. Chatterjee SK. Scar endometriosis: a clinicopathologic study of 17 cases. Obstet Gynecol 1980; 56: 81–84. 6. Patterson GK, Winburn GB. Abdominal wall endometriomas: report of eight cases. Am Surg 1999; 65: 36–39. 7. Francica G, Giardiello C, Angelone G, Cristiano S, Finelli R, Tramntano G. Abdominal wall endometriomas near Cesarean delivery scars: sonographc and Doppler findings in a series of 12 patients. J Ultrasound Med 2003; 22: 1041–1047. 8. Francica G, Scarano F, Angelone G, Giardiello C. Endometriomas in the region of a scar from Cesarean section: sonographic appearance and clinical presentation vary with the size of the lesion. J Clin Ultrasound 2009; 37: 215–220.

SUPPORTING INFORMATION ON THE INTERNET The following supporting information may be found in the online version of this article: Videoclip S1 Nodular abdominal wall endometriosis in a scarless abdomen: transverse scan of the abdominal wall at a level between the transverse suprapubic line and the umbilicus, in a woman without any previous abdominal surgery. Note the presence of a 3-cm hypoechoic solid nodule with a hyperechoic rim deeply extending into the subcutaneous fatty tissue towards the abdominal fascia.

Copyright  2012 ISUOG. Published by John Wiley & Sons, Ltd.

Ultrasound Obstet Gynecol 2012; 39: 336–340.