British Journal of Radiology (2006) 79, 303-307
© 2006 British Institute of Radiology
doi: 10.1259/bjr/70121266
Prevalence and distribution of adnexal findings suggesting endometriosis in patients with MR diagnosis of adenomyosis
T T Zacharia, MD
and
M J O'Neill, MD
Abdominal Imaging and Interventional Radiology, Massachusetts General Hospital, Harvard Medical School, USA
Correspondence: T Thomas Zacharia, MD, #4225, Elmhurst Ave, Ithaca Street, Queens, New York City, NY 11373, USA.
 |
Abstract
|
|---|
The purpose of this investigation was to establish the prevalence and distribution of MR findings associated with pelvic endometriosis in patients with a MRI diagnosis of adenomyosis. Retrospective review of the pelvic MRI in 59 patients (age 32–54 years, mean 42 years) who met strict MRI criteria for adenomyosis was performed. T1 weighted fat saturated and T2 weighted images in these patients were reviewed for the presence or absence of T1 bright signal suggesting endometriosis in any of nine locations (uterine serosa, right and left ovary, right and left fallopian tube, right and left broad ligament, and right and left pelvic side wall). 20 (20/59) patients (34%), showed characteristic MRI features associated with endometriosis. A total of 54 sites of involvement were identified (uterine serosa n = 17, ovaries n = 14, broad ligaments n = 10, fallopian tubes n = 8, pelvic side walls n = 5) in 20 patients with an average of 2.7 sites per patient. Implants (n = 43) were more common than endometriomas (n = 11). Endometriomas occurred most often in the ovaries (ovaries n = 9, broad ligament n = 2) whereas implants were seen on all locations (uterine serosa n = 17, ovaries n = 5, broad ligaments n = 8, fallopian tubes n = 8, pelvic side walls n = 5). One third of patients with adenomyosis diagnosed by MRI also had MRI findings of endometriosis, with serosal implants being the most common finding. Imaging protocols should routinely include T1 weighted fat saturated imaging sequences in order to detect coexistent endometriosis in patents undergoing pelvic MRI for the diagnosis of adenomyosis.
|
Introduction
|
|---|
Adenomyosis and endometriosis are gynaecological disorders with similar pathological appearance and proposed aetiologies. We sought to define whether these disease entities coexist.
Recent studies have shown high sensitivity and specificity of MRI in the diagnosis of endometriosis [1, 2]. The detection and characterization of endometrioma is relatively easier when compared with detection of serosal endometriotic implants. A diagnosis of serosal implants can be easily missed during routine pelvic MRI evaluation.
Some investigations performed recently have shown MRI to have very high sensitivity and specificity in the diagnosis of adenomyosis [3, 4]. MRI has shown added value in the evaluation of associated diseases like uterine leiomyomas when compared with transvaginal ultrasound (TVUS).
Similar theories of abnormal migration and metaplasia [5–7] are postulated in the aetiology of both adenomyosis and endometriosis. The most commonly proposed aetiology for adenomyosis is direct migration of endometrial tissue into the myometrium. This mechanism of direct extension is similar to the most widely accepted theory for the aetiology of endometriosis, i.e. reflux and retrograde extension of endometrial tissue into the peritoneal cavity through the fallopian tube.
To date, no large series of patients with both disease entities has been reported. The purpose of this investigation was to establish the prevalence and distribution of MR findings associated with pelvic endometriosis in patients with a MRI diagnosis of adenomyosis.
|
Materials and methods
|
|---|
Patients
Retrospective review of an MRI database from 1 January 2000 to 6 January 2003 identified 63 patients (aged 32–54 years, mean 42 years) in whom the diagnosis of adenomyosis was included in the differential diagnosis. Indications for MRI included menstrual irregularities (like dysmenorrhoea, menorrhagia), pelvic pain, infertility and clinically palpable uterus. IRB (Institutional Review Board) approval and informed consent was obtained in all patients. Retrospective review by a subspecialty trained women's imaging radiologist of the pelvic MRI exams in these patients found 59 patients who met strict MRI criteria (Reinhold's criteria) for adenomyosis (junctional zone (JZ) >12 mm and sub endometrial T2 bright foci within abnormal JZ) [8] (Figure 1a
). These 59 patients served as the investigational study group. 20 (20/59) patients had severe symptoms. Hysterectomy was performed in all these 20 patients. To serve as a control group there should ideally have been a review of MR exams performed for the clinical indication of endometriosis, to evaluate co-existence of adenomyosis. However, a converse study (detection of adenomyosis in patients with MR diagnosis of endometriosis) was not performed as part of our investigation.
View larger version (107K):
[in this window]
[in a new window]
|
Figure 1. 35-year old woman with posterior wall adenomyoma and coexisting right ovarian endometrioma.(a) Sagittal T2 weighted MR image (repetition time (TR) 5800/echo time (TE) 100) shows classical findings of adenomyosis involving the posterior wall of the uterus. Ill defined adenomyoma with T2 bright focus within (arrow). (b) Coronal T2 weighted MR image (TR 5000/TE 125) shows right ovarian endometrioma with T2 shading (arrow). (c) Axial T1 weighted fat saturated MR image (TR 150/TE 1.8) shows very hyperintense right ovarian endometrioma (arrow).
|
|
MRI protocols
MRI was performed with 1.5 T signa scanners (General Electric Medical Systems, Milwaukee, WI). T2 weighted images were acquired with 4 mm slices with 1 mm spacing in the sagittal, coronal, and axial planes relative to the orientation of uterine cavity, using fast (turbo) spin echo sequences in all three planes. T1 weighted spin echo and gradient echo images were obtained both with and without fat saturation in all 59 patients. High resolution surface coils were used for data acquisition and exams were completed in 30–45 min.
Image analysis
T1 weighted fat saturated and T2 weighted images in these patients were then reviewed for the presence or absence of T1 bright/T2 dark signal suggesting endometriosis in any of nine anatomical locations. Endometrial implant has a varied appearance depending on the age of associated blood products. Diagnosis of implant was made when hyperintense lesions were detected on fat saturated T1 weighted imaging. The diagnosis of endometrial implant was made after ruling out artefacts of motion, incomplete fat saturation and bowel contents. Lesions larger that 10 mm were considered to represent endometriomas (Figure 1b,c
) and lesions less than 10 mm were classified as implants [9]. The sites (n = 9) evaluated included uterine serosa, bilateral ovaries, bilateral fallopian tubes, bilateral broad ligaments and pelvic side walls. Fallopian tubes were distinguished from broad ligament on coronal MR images as thin tubular structures.
Data analysis
The reviewed data were collected on a Microsoft Excel (Redmond, DC) spreadsheet. The data entered included patient age, sex, MRI findings and sites of involvement. The percentage of patients with coexistent endometriosis and adenomyosis was calculated.
|
Results
|
|---|
Approximately one third of patients diagnosed to have adenomyosis by MRI were found to have coexisting endometriosis. 20 (20/59) patients with adenomyosis (34%, 95% confidence interval (CI) 22–47%)) had coexisting endometriosis (Figure 2).
View larger version (88K):
[in this window]
[in a new window]
|
Figure 2. 38-year-old woman with adenomyosis and serosal endometriosis.(a) Sagittal T2 weighted MR image (repetition time (TR) 5800/echo time (TE) 100) shows junctional zone thickening (arrow) consistent with adenomyosis. (b) Sagittal gradient echo T1 weighted fat saturated MR image (TR 150/TE 1.8) shows hyperintense focus (arrow) suggestive of serosal implant.
|
|
Uterine serosa and ovaries were the most common sites of involvement in our series (Table 1). Single site involvement was noted in two patients and multiple sites were involved in 18 patients. Extra ovarian involvement was noted in all patients (n = 20).
Sites of endometriosis
The various sites of involvement studied were uterine serosa, bilateral ovaries, broad ligaments, fallopian tubes and the pelvic wall. A total of 54 sites of involvement were identified (uterine serosa n = 17/54, ovaries n = 14/54, broad ligaments n = 10/54, fallopian tubes n = 8/54, pelvic side walls n = 5/54) in 20 patients with an average of 2.7 sites per patient.
Implants
Implants (n = 43) were more common than endometriomas (n = 11) and formed 79% of the lesions. Implants (lesions less than 10 mm) were most frequently seen over the uterine surface (Figure 2b). Most of these lesions involving uterine serosa were more than 2 mm thick (Figure 1), but occasionally these lesions were subtle and barely perceptible (Figure 2). Implants were seen on all locations (uterine serosa n = 17, ovaries n = 5, broad ligaments n = 8, fallopian tubes n = 8, pelvic side walls n = 5).
Endometriomas
Endometriomas were less common than serosal implants (n = 11) (20.4%). Endometriomas occurred most often in the ovaries (ovaries n = 9, broad ligament n = 2). "T2 shading" was observed in some of these lesions (Figure 1b). The diagnosis was obvious also on T1 fat saturated sequence.
Histopathology
Histopathological confirmation was obtained in 20 patients (20/59) (34%) who underwent hysterectomy. All of them had findings correlating with MRI. Seven (7/20) patients had coexisting adenomyosis and endometriosis. The presence of multiple patchy or diffuse areas of endometrial glands and/or stroma associated with surrounding muscle hypertrophy located more than halfway from the endometrial-myometrial junction was considered diagnostic of adenomyosis on histopathology. Endometriosis was diagnosed when endometrial tissue (endometrial gland and stroma) was found at histopathological examination of resected specimen in uterine serosa, ovaries, broad ligaments, fallopian tubes or pelvic sidewalls.
|
Discussion
|
|---|
Common aetiological factors like abnormal migration, metaplasia and catalase are implicated in the pathogenesis of both endometriosis and adenomyosis [5–7]. Sampson's theory of retrograde menstruation is the most accepted theory postulated in the pathogenesis of endometriosis and has been supported by convincing circumstantial and experimental evidence [5]. Metaplasia of serosal epithelium is another postulated theory in the aetiology of endometriosis [7].
Adenomyosis is a fairly frequent disorder in adult women characterized by infiltration of endometrial glands and stroma within the myometrium of the uterus. Recent experimental data have supported the hypothesis that adenomyosis represents a condition of the uterine body in which the stromal cells have a primary pathogenetic role, although some contribution of accelerated epithelial down growth cannot be entirely excluded [7]. This migration of endometrial epithelium to the inner myometrium can act as an irritant focus and cause smooth muscle hyperplasia.
There are a number of clinical and pathological studies, which throw light on the coexistence of both the diseases [1, 7, 10]. However, no imaging study exists which has studied the coexistence of adenomyosis and endometriosis. We sought to evaluate the patients with adenomyosis who had coexisting serosal or ovarian endometriosis on MRI.
MRI findings in endometriosis has been well documented in the literature. The endometriotic lesions are typically hyperintense on T1 weighted fat saturated sequence with "shading" on T2 weighted imaging [11–13]. Identification of these lesions by MRI relies on detection of haemorrhagic lesions.
Signal characteristics vary according to the age of the haemorrhage. Typically, these lesions appear hyperintense on T1 weighted spin echo images and hypointense (shading) on T2 weighted fast spin echo images (T2WFSE) due to the presence of deoxyhaemoglobin and met haemoglobin. Acute haemorrhage occasionally appears hypointense on T1WSE and T2WFSE sequences. Old haemorrhage occasionally appears hyperintense on T1WSE and T2WFSE images. T1 weighted fat saturation sequence was performed in all our patients. The addition of fat saturated T1 weighted imaging improved diagnostic accuracy in the evaluation of both endometriomas and peritoneal disease by increasing lesion conspicuity and differentiating lipid containing ovarian masses from those containing blood. Recent studies have shown that administration of gadolinium will enhance detection of deep endometriotic implants.
The coexistence of endometriosis and adenomyosis has got significant clinical implications. The purpose of this investigation is to prove the association between endometriosis and adenomyosis. The observation that one third of patients with adenomyosis have coexisting endometriosis should alert the radiologist to look for endometriosis, which is quite often subtle, represented by uterine serosal implants. The pelvic sidewall implants can also be easily missed. In patients presenting with infertility, the coexistence of these two conditions can result in treatment failures if both the diseases are not detected and treated. Hysterectomies can be avoided if endometriosis can be diagnosed and treated by drugs like progestin, GnRH (gonadotrophin releasing hormone) analogues and Danazol [14, 15]. If symptomatic relief is attained in this group of patients after medical management of endometriosis, hysterectomy may not be needed to treat adenomyosis.
Our findings have important implications in deciding the MRI protocols too. Fat saturated T1 weighted imaging should be routinely included in the pelvic MRI protocol done for patients undergoing evaluation for adenomyosis.
The most important drawback of our study is the lack of pathological evidence in patients who did not have the severe symptoms to undergo hysterectomy. Although MRI has high specificity in the diagnosis of both these disease entities, certain centres depend upon deep myometrial biopsy for the diagnosis of adenomyosis and laparoscopic biopsy for the diagnosis of endometriosis [16]. Another limitation is the absence of a control group. There should have been a converse study (a review of MR exams performed for endometriosis, to evaluate associated adenomyosis) and this was not performed as part of our investigation. Without rigorous statistics, this association will be difficult to prove absolutely. However, similar pathogenesis, histological appearance and presentation in a similar age group suggest that the association might be real.
|
Conclusion
|
|---|
MRI evidence of endometriosis was seen in one third of patients diagnosed to have adenomyosis on MRI. Serosal implants especially on the uterine surface were the most common observation. Endometriomas were also commonly noted. Imaging protocols should routinely include T1 weighted fat saturated imaging sequences in order to detect coexistent endometriosis in patents undergoing pelvic MRI for the diagnosis of adenomyosis.
|
Footnotes
|
|---|
Presented at a meeting: 104th ARRS (American Roentgen Ray Society) meeting, Miami, Florida, May 2004. 
Received for publication May 19, 2005.
Revision received July 17, 2005.
Accepted for publication August 12, 2005.
|
References
|
|---|
- Dumontier I, Roseau G, Vincent B, et al. Comparison of endoscopic ultrasound and magnetic resonance imaging in severe pelvic endometriosis. Gastroenterol Clin Biol 2000;24:1197–204.[Medline]
- Stratton P, Winkel C, Premkumar A, et al. Diagnostic accuracy of laparoscopy, magnetic resonance imaging, and histopathologic examination for the detection of endometriosis. Fertil Steril 2003;79:1078–85.[CrossRef][Medline]
- Bazot M, Cortez A, Darai E, Rouger J, Chopier J, Antoine JM, et al. Ultrasonography compared with magnetic resonance imaging for the diagnosis of adenomyosis: correlation with histopathology. Hum Reprod 2001;16:2427–33.[Abstract/Free Full Text]
- Dueholm M, Lundorf E, Hansen ES, Sorensen JS, Ledertoug S, Olesen F. Magnetic resonance imaging and transvaginal ultrasonography for the diagnosis of adenomyosis. Fertil Steril 2001;76:588–94.[CrossRef][Medline]
- Sampson JA. Benign and malignant endometrial implants in the peritoneal cavity, and their relation to certain ovarian tumors. Surg Gynecol Obstet 1924;38:287–311.
- Ota H, Igarashi S, Sato N, Tanaka H, Tanaka T. Involvement of catalase in the endometrium of patients with endometriosis and adenomyosis. Fertil Steril 2002;78:804–9.[CrossRef][Medline]
- Starzinski-Powitz A, Zeitvogel A, Schreiner A, Baumann R. In search of pathogenic mechanisms in endometriosis: the challenge for molecular cell biology. Curr Mol Med 2001;1:655–64.[CrossRef][Medline]
- Reinhold C, McCarthy S, Bret PM, Mehio A, Atri M, Zakarian R, et al. Diffuse adenomyosis: comparison of endovaginal US and MR imaging with histopathologic correlation. Radiology 1996;199:151–8.[Abstract/Free Full Text]
- Manfredi R, Valentini AL. Magnetic resonance imaging of pelvic endometriosis. Rays 1998;23:702–8.[Medline]
- Sharpe-Timms KL. Endometrial anomalies in women with endometriosis. Ann NY Acad Sci 2001;943:131–47.[Medline]
- Gougoutas CA, Siegelman ES, Hunt J, Outwater EK. Pelvic endometriosis: various manifestations and MR imaging findings. AJR Am J Roentgenol 2000;175:353–8.[Free Full Text]
- Umaria N, Olliff JF. Imaging features of pelvic endometriosis. Br J Radiol 2001;74:556–62.[Abstract/Free Full Text]
- Glastonbury CM. The shading sign. Radiology 2002;224:199–201.[Free Full Text]
- Ylanen K, Laatikainen T, Lahteenmaki P, Moo-Young AJ. Subdermal progestin implant (Nestorone) in the treatment of endometriosis: clinical response to various doses. Acta Obstet Gynecol Scand 2003;82:167–72.[CrossRef][Medline]
- Rotondi M, Labriola D, Rotondi M, Ammaturo FP, Amato G, Carella C, et al. Depot leuprorelin acetate versus danazol in the treatment of infertile women with symptomatic endometriosis. Eur J Gynaecol Oncol 2002;23:523–6.[Medline]
- McLucas B, Perrella R, Adler L. Embolization for the treatment of adenomyosis. AJR Am J Roentgenol 2002;178:1028–9.[Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
S Ghattamaneni, M J Weston, and J A Spencer
Imaging in endometriosis
Imaging,
December 1, 2007;
19(4):
345 - 368.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
