Multifocal polypoid endometriosis presenting as huge pelvic masses causing deep vein thrombosis.

Multifocal Polypoid Endometriosis Presenting as Huge Pelvic Masses Causing Deep Vein Thrombosis Full access Lisa A. Laird From the Department of Pathology (Dr Laird), Division of Gynecologic Oncology, Department of Obstetrics and Gynecology (Dr Hofmann), New Britain General Hospital, New Britain, Conn; and the Department of Obstetrics and Gynecology, University of Connecticut Health Center, Farmington (Dr Omrani) From the Department of Pathology (Dr Laird), Division of Gynecologic Oncology, Department of Obstetrics and Gynecology (Dr Hofmann), New Britain General Hospital, New Britain, Conn; and the Department of Obstetrics and Gynecology, University of Connecticut Health Center, Farmington (Dr Omrani) From the Department of Pathology (Dr Laird), Division of Gynecologic Oncology, Department of Obstetrics and Gynecology (Dr Hofmann), New Britain General Hospital, New Britain, Conn; and the Department of Obstetrics and Gynecology, University of Connecticut Health Center, Farmington (Dr Omrani) We describe a case of multifocal polypoid endometriosis presenting with advanced bulky disease at a variety of pelvic sites. The extent of tumor and clinical features such as vaginal bleeding and pulmonary embolus were suggestive of a malignant process. Histopathology demonstrated glands that were neither crowded nor complex, with intervening fibromatous stroma that contained occasional endometrial stromal cells. These features were consistent with the newly described condition of polypoid endometriosis. Despite the endometrioid appearance of this tumor, there was florid ciliary cell change. An association has been suggested between polypoid endometriosis and prior tamoxifen use, although this patient had no history of prior hormone use. Polypoid endometriosis is a recently described entity, which can present as tumor masses simulating the clinical course of neoplastic or even malignant conditions.1–4 The lack of histologically atypical features distinguishes it from these other neoplasms. Differential diagnoses of tumors with endometrial glands in a fibromatous stroma include adenofibromas, borderline tumors, and even adenosarcomas. The presence of abundant cilia in an endometrioid ovarian tumor also raises the possibility of an adenofibromatous ciliated-cell tumor of the ovary, a lesion with similar histology but lacking endometrial stroma.5 REPORT OF A CASE An 80-year-old, gravida 0 black woman with a long history of schizophrenia had been residing in a state facility. She was transferred to our hospital because of painful swelling in her left leg and sudden onset of dyspnea. An abdominal mass had been present for an indeterminate duration. Vaginal bleeding had also been present for an unspecified period of time. The painful swelling of her leg had been present for several months. A venous Doppler study confirmed venous thrombosis of the deep femoral vein. A spiral computed tomographic study of the chest showed an embolus in the left main pulmonary artery. Computed tomography also detailed large masses involving the uterus, left ovary, and left pelvic nodes. The images revealed a complex, predominately cystic mass in the left adnexa. There was also gross enlargement of the uterus with fluid and debris in the endometrial cavity. Bulky left pelvic adenopathy was present in the region of the obturator nodes. Aggregate dimension of these masses was greater than 30 cm (Figures 1 and 2). Serum CA 125 assay was measured to be 323 U/mL (normal, <36 U/ mL). The patient was treated with heparin for her deep femoral vein thrombosis and pulmonary artery embolus. After discussion with her conservator, a decision was reached to attempt surgical removal of her tumors. The anticoagulation was reversed, and a Greenfield filter was placed in the inferior vena cava prior to performing a laparotomy. At the time of the operation some ascites was present in the abdominal cavity, but there was no overt evidence of peritoneal metastases. The left ovary was replaced by a predominately cystic mass that measured 17 cm. Large solid tumors were discovered throughout the uterus and cervix, connected to similar masses in the parametrial and pararectal spaces. The left external iliac and obturator lymph nodes were grossly enlarged and were similar in texture to these other lesions. The other pelvic node groups appeared normal. Polypoid tumors extended from the cervical canal and arose directly from the vagina. A radical hysterectomy, bilateral salpingo-oophorectomy, and pelvic lymph node dissection were able to remove all of these masses. Postoperatively, the patient's recovery was rapid and resulted in complete resolution of her leg swelling and of all of her respiratory abnormalities. She continues to do well 2 years after her surgery. CA 125 serum assay was reduced to 7 U/mL. PATHOLOGIC FINDINGS Multiple cystic and solid nodules ranging from 0.7 to 5.4 cm in greatest dimension were identified within the left ovary, the periureteral tissues, the parametrium, the pararectal space, and the obturator nodes. The masses all displayed a similar appearance, often polypoid, with a red-brown gelatinous cut surface containing multiple cysts. No papillary or necrotic areas were identified. Within the uterus was a similar mass at the junction of the ectocervix and the vaginal fornix. In addition, the endometrium contained a 6.0-cm, multinodular, lobulated polypoid mass, which on sectioning displayed a white, solid cut surface. It did not appear to extend into the myometrium. Microscopically, the extrauterine masses were similar and were composed of various-sized glands lined by a single layer of cuboidal to columnar cells. In general, the glands were not crowded and no architectural complexity was found. These glands were endometrioid in appearance and displayed no significant cytologic atypia (Figure 3). Mitotic rate was less than 1 per 50 high-power fields. Tubal metaplasia with abundant ciliary change was present, but no mucin was identified (Figure 4). The intervening stroma contained numerous oval to spindle-shaped cells, consistent either with fibroblasts or with endometrial stromal cells (Figure 5). CD10 staining confirmed the presence of endometrial stromal cells (Figure 6). We did not note areas of periglandular condensation of stromal cells, as is seen in adenosarcoma. Individually, the lesions were well demarcated from the surrounding normal tissue and lacked infiltrating borders. The polypoid endometrial lesion differed somewhat. A portion of the polyp had the appearance of an atypical adenomyoma, with crowded glands separated by smooth muscle. In other areas, the glands assumed a back-to-back appearance with no intervening stromal tissue, compatible with a well-differentiated adenocarcinoma arising within a background of simple hyperplasia with atypia. No underlying invasion into the myometrium was found. COMMENT The differential diagnosis of multifocal tumors with mullerian epithelium and fibrous stroma includes a wide array of malignant, benign, and even nonneoplastic lesions. In the present case, the epithelial component was notable for the presence of frequent cilia, an uncommon finding in endometrioid neoplasms. The stromal component was shown to have both fibromatous and endometrial stromal cell components. This combination of findings allowed us to classify this tumor as an example of polypoid endometriosis, an entity only recently described.1–4 Several other allied conditions were given careful consideration in the differential diagnosis, and it is important to review the relationship of our case to these entities. Benign endometrioid adenofibroma of the ovary can present a histologic picture that is similar to our case, but it has not been reported in such a disseminated fashion. The widespread distribution of the current case would be more consistent with a borderline endometrioid tumor. Most borderline endometrioid tumors have a fibromatous stroma, as did our case, but this diagnosis requires at least atypical if not cytologically malignant epithelial glands.6 Very rare, ciliated variants of ovarian adenofibroma and borderline tumor were recently described by Eichhorn and Scully,5 and these were major elements in our differential diagnosis. The presence of endometrial stromal cells and absence of epithelial atypia in our case favor polypoid endometriosis rather than a ciliated ovarian tumor. Endometrioid adenocarcinoma is likewise excluded, owing to the lack of cytologic atypia. Endometrial stromal tumors, especially low-grade endometrial stromal sarcomas, can contain glandular differentiation, which is occasionally pronounced. The glands are almost always of endometrioid type and appear benign, although rare cases of atypical or carcinomatous glands have been described.7 The predominant stromal cell component is endometrial, with mild nuclear pleomorphism and increased mitotic activity.8 The distribution of the tumors may be multifocal and polypoid, as the current case, but the predominance of a spindled fibroblastic stroma precludes this categorization. Another tumor composed of epithelial and mesenchymal elements is adenosarcoma, characterized as containing benign epithelial but malignant stromal components. While most adenosarcomas occur in the uterus, other extrauterine primary sites have been reported.8 Metastases are invariably composed predominately or exclusively of the sarcomatous element, and as such the widespread presence of benign mullerian tissue at multiple extrauterine sites eliminates the possibility of metastatic sarcoma, although it could be seen in cases of direct extension or multifocal origin within diffuse endometriosis. The nuclear features of the stroma may resemble endometrial stroma cells or fibroblasts, but should show at least mild atypia and hypercellularity, especially in periglandular areas.89 These latter features were absent in our case. The most common manifestation of multifocal mullerian epithelium is endometriosis. The age, distribution of disease, and clinical presentation of this patient are wholly unlike the usual setting of conventional endometriosis. However, polypoid endometriosis has been described in patients aged 23 to 89 years and includes cases with solitary or multiple polypoid masses in the pelvis and abdominal cavity.12 In a recent review, Dadmanesh et al1 described 15 women, aged 23 to 78 years, with polypoid endometriosis who presented with pelvic masses (10 cases), colon obstruction (2 cases), cul-de-sac mass (1 case), periadrenal mass (1 case), and umbilical mass (l case). In 10 of the patients these masses were found in multiple sites. The authors viewed this entity as a manifestation of endometriosis resembling a neoplastic process. The tumors measured up to 12 cm and were sometimes friable, hemorrhagic, or cystic. Four of the 15 patients had a history of estrogen therapy, but exogenous hormone therapy was not reported to be a feature in the remaining cases. Schlesinger and Silverberg2 identified the antecedent use of tamoxifen in 12 of 17 reported cases of polypoid endometriosis. It is unclear from the literature whether tamoxifen is in fact a principal cause of this entity, since these cases were all individual case reports, whereas patients in the large series of patients by Dadmanesh et al1 did not seem to share this feature. Less commonly, the condition has been described in association with withdrawal of gonadotropin-releasing hormone agonist.10 In the review by Schlesinger and Silverberg,2 2 of 12 patients with a history of tamoxifen use antecedent to their polypoid endometriosis had an associated endometrial adenocarcinoma. We noted this association in our patient despite the absence of prior tamoxifen use. Schlesinger and Silverberg2 referenced 3 cases presenting for ureteral obstruction and identified sites of disease throughout the pelvis, vagina, omentum, and periureteral tissues. Our patient had tumor involving these many sites in addition to involvement of pelvic nodes and the pararectal space. These areas of spread occasioned both obstruction of the ureter and deep vein thrombosis. Endometrial glands in polypoid endometriosis have been described with variable appearances, including proliferative and secretory phases, as well as hyperplastic changes (simple and complex, with and without atypia). Many forms of metaplasia have also been described within the glandular elements, including squamous, mucinous, and focal tubal metaplasias.1210 However, our case was unusual in that it had extensive ciliated-cell change suggestive of the pattern reported in endometrioid ciliated-cell tumor of the ovary.5 Nonetheless, based on the distribution of the disease and the stromal appearance, we do feel our case is an example of polypoid endometriosis. As more cases of this entity are described, we hope to see reports about postoperative treatment and patterns of recurrence. We plan to perform clinical examinations, CA 125 assays, and periodic computed tomographic scanning to determine whether our patient develops recurrent disease. Consideration will be given for hormonal treatment in the event of a recurrence. Copyright: College of American Pathologists 2004 Figure 1. Large masses are seen throughout the uterus (black arrow) and in the left pelvic nodes (white arrow). Figure 2. Large, multiloculated cystic tumor from left ovary. Arrow shows left hydroureter. Figure 3. Somewhat dilated endometrioid glands (hematoxylin-eosin, original magnification ×100). Figure 4. Focus of tubal metaplasia with abundant cilia (arrow) (hematoxylin-eosin, original magnification ×400). Figure 5. Focus of endometrial stromal cells (arrow) (hematoxylin-eosin, original magnification ×200). Figure 6. Diffuse strong CD10 staining of endometrial stromal cells (original magnification ×100) Contributor Notes Reprints: Lisa Laird, MD, Department of Pathology, New Britain General Hospital, 100 Grand St, New Britain, CT 06050