{"paper_id":"7e698c5c-d6ca-40ec-9ddf-3e9299446a20","body_text":"Primary Peritoneal Anaplastic Giant Cell Carcinoma: Case Report of an Unusual and Highly Malignant Müllerian Neoplasm\nFull access\nXian Lu\nFrom the Department of Pathology and Laboratory Medicine, Women & Infants Hospital of Rhode Island, Warren Alpert Medical School of Brown University, Providence, RI. Dr Lu is now with the Department of Clinical Pathology, The First Hospital, Shanxi Medical University, Taiyuan, Shanxi Province, People's Republic of China\nFrom the Department of Pathology and Laboratory Medicine, Women & Infants Hospital of Rhode Island, Warren Alpert Medical School of Brown University, Providence, RI. Dr Lu is now with the Department of Clinical Pathology, The First Hospital, Shanxi Medical University, Taiyuan, Shanxi Province, People's Republic of China\nFrom the Department of Pathology and Laboratory Medicine, Women & Infants Hospital of Rhode Island, Warren Alpert Medical School of Brown University, Providence, RI. Dr Lu is now with the Department of Clinical Pathology, The First Hospital, Shanxi Medical University, Taiyuan, Shanxi Province, People's Republic of China\nFrom the Department of Pathology and Laboratory Medicine, Women & Infants Hospital of Rhode Island, Warren Alpert Medical School of Brown University, Providence, RI. Dr Lu is now with the Department of Clinical Pathology, The First Hospital, Shanxi Medical University, Taiyuan, Shanxi Province, People's Republic of China\nFrom the Department of Pathology and Laboratory Medicine, Women & Infants Hospital of Rhode Island, Warren Alpert Medical School of Brown University, Providence, RI. Dr Lu is now with the Department of Clinical Pathology, The First Hospital, Shanxi Medical University, Taiyuan, Shanxi Province, People's Republic of China\nSearch for other papers by Margaret M. Steinhoff in\nFrom the Department of Pathology and Laboratory Medicine, Women & Infants Hospital of Rhode Island, Warren Alpert Medical School of Brown University, Providence, RI. Dr Lu is now with the Department of Clinical Pathology, The First Hospital, Shanxi Medical University, Taiyuan, Shanxi Province, People's Republic of China\nVirtually all primary peritoneal carcinomas (PPCs) are of serous papillary type. We report an unusual histologic type of PPC composed of anaplastic giant cells, which exhibited an aggressive clinical course. A 72-year-old woman presented with lower abdominal pain. Computed tomography showed a diffuse omental thickening. The patient underwent an exploratory laparotomy with omentectomy, total hysterectomy, bilateral salpingo-oophorectomy, and appendectomy. Pathologic examination revealed extensive omental replacement by tumor but only superficial surface cortical involvement of both ovaries, a disease distribution consistent with a typical müllerian-derived PPC. However, this neoplasm was composed of diffuse anaplastic tumor giant cells, rather than serous carcinoma, which is the usual histologic type encountered in PPC. The patient died within 1 month after surgery. We report this unusual histologic variant of PPC to raise awareness that anaplastic giant cell carcinoma may arise in the pelvic peritoneum as a primary tumor.\nPrimary peritoneal carcinoma (PPC) is a rare tumor and tends to occur in elderly women.1 Like its ovarian counterpart, the tumor often presents with abdominal distention, pain, and pressure.12 More than 90% of PPCs are papillary serous carcinomas.1–3 Other much less common types of PPC include clear cell carcinoma, endometrioid carcinoma, mucinous carcinoma, and carcinosarcoma. Primary anaplastic giant cell carcinomas have been described in a number of mostly nongynecologic sites and occasionally in the ovaries,4 but they have not been reported in the pelvic peritoneum. We report an unusual histopathologic type of primary peritoneal carcinoma comprising diffuse anaplastic tumor giant cells.\nREPORT OF A CASE\nThe patient was a previously healthy 72-year-old woman who presented with a 5-week history of lower abdominal pain and bloating. Physical examination showed abdominal distention and probable ascites. Computed tomography revealed diffuse omental thickening and a small amount of free pelvic fluid. Serum CA 125 and carcinoembryonic antigen (CEA) levels were elevated to 260 U/mL and 4.2 ng/mL, respectively. Exploratory laparotomy showed extensive thickening of the omentum, small tumor nodules in the appendix and bilateral ureters, and roughness of the cortices of both ovaries. Following a frozen section diagnosis of high-grade carcinoma, the patient underwent a total hysterectomy, bilateral salpingo-oophorectomy, and an appendectomy; the surgical procedure removed all the visible tumors, except that involving both ureters.\nGross examination revealed the omentum (25.0 × 7.0 × 3.0 cm) to be markedly thickened and replaced by diffuse firm nodules. The cut surfaces of the nodules were granular and varied from yellow-white to hemorrhagic. The uterus and the bilateral fallopian tubes, except for focal serosal roughness, were unremarkable. The right and left ovaries measured 2.5 × 2.0 × 0.9 cm and 2.5 × 2.0 × 0.6 cm, respectively. Both ovaries showed focal surface adhesions, but their parenchymal cut surfaces were unremarkable. Focal thickening of the right paratubal tissue was noted, as was a 2.0-cm tumor nodule on the serosal surface of the appendix.\nRoutine hematoxylin-eosin stains showed complete replacement of the omentum by tumor cells arranged in solid sheets, with intervening benign fibroblastic and inflammatory cells (Figure, A and B). One 2-mm focus, identified in one section from 8 blocks of the omentum examined, exhibited irregular slitlike spaces, consistent with serous differentiation (Figure, C). The most striking histologic feature, however, was the presence of diffusely infiltrative tumor giant cells, whose nuclei were up to 40 times larger than those of the adjacent benign cells (Figure, B). The nuclei of the giant cells were large, hyperchromatic, and bizarre, with frequent “popcorn” nuclear forms, but multinucleation (defined as separate nuclei in a given cell) was not a prominent feature. Because of their large size and hyperchromasia, many neoplastic nuclei simulated microcalcifications at scanning magnification, but true psammomatous calcifications were scarce (Figure, A). The tumor exhibited a high mitotic index, in the range of 10 to 15 per 10 high-power fields. The abundant cytoplasm was eosinophilic to amphophilic and contained variably sized vacuoles.\nMicroscopic examination disclosed similar tumors in the right paratubal tissue, on the cortical surfaces of both ovaries without significant parenchymal invasion (Figure, D), and on the serosal surfaces of both fallopian tubes, the uterus, and the appendix.\nThe tumor cells were strongly immunopositive for AE1/AE3 (Figure, E); CA 125 (Figure, F); and B72.3, cytokeratin (CK) 7, WT-1, and p53; were focally positive for vimentin, CEA, and S100 protein; and were negative for human chorionic gonadotropin (HCG), desmin, HMB-45, Melan A, smooth muscle actin (SMA), CK20, CDX2, gross cystic disease fluid protein 15 (GCDFP-15), thyroid transcription factor 1 (TTF-1), calretinin, estrogen receptors, progesterone receptors, CD45, and placental-like alkaline phosphatase (PLAP). Some tumor cells exhibited mucicarmine positivity. CD68 positivity was seen in the adjacent benign-appearing cells (likely histiocytes) but not in the neoplastic cells.\nBased on its location and distribution, histology, and immunophenotype, the tumor was categorized as a primary peritoneal anaplastic giant cell carcinoma.\nPostoperatively, the patient received cisplatin-based chemotherapy. However, she developed hydronephrosis and urinary toxemia secondary to tumor involvement of both ureters. The patient experienced a rapid deterioration of her general condition and died within 1 month after surgery.\nCOMMENT\nTypical primary peritoneal carcinoma often presents with disseminated intraperitoneal carcinomatosis in the face of relatively normal-sized ovaries, and it is histologically identical to other müllerian-derived serous carcinomas, such as those of the ovary and endometrium. The diagnosis of PPC is typically made by exclusion after both operative assessment and pathologic examination. The Gynecologic Oncology Group has agreed upon several criteria for the diagnosis of PPC5. (1) Both ovaries are either normal in size or enlarged by a benign process. (2) In the judgment of the surgeon and the pathologist, the bulk of the tumor is in the peritoneum, and the extent of tumor involvement at one or more extraovarian sites is greater than on the surface of either ovary. (3) Microscopic examination of the ovaries reveals either (a) no tumor, or one of the following scenarios: (b) tumor is confined to the surface epithelium, with no evidence of cortical invasion, (c) tumor involves the ovarian surface and underlying cortical stroma but is smaller than 5 mm, or (d) tumor measures less than 5 mm in greatest dimension within the ovarian substance with or without surface involvement. (4) The histologic characteristics of the tumor are similar or identical to those of ovarian serous papillary adenocarcinomas.\nIn our patient, extensive tumor involved the omentum, and there was focal serosal involvement of both ovaries and fallopian tubes, the uterus, and the appendix. The ovaries were normal in size and showed no cortical involvement by tumor. The disease distribution was consistent with a primary peritoneal carcinoma, but the histologic type was different from that of the vast majority of PPCs. The latter types are characterized by variably differentiated, but often high-grade, serous carcinoma, sometimes but not invariably showing at least a papillary component.1–3 In our patient, virtually the entire neoplasm was composed of solid sheets of diffuse anaplastic giant cells.\nOnly one other case of primary peritoneal carcinoma with a component of diffusely distributed giant cells has been described6; however, it contained numerous multinucleated giant cells that were positive for HCG and likely represented syncytiotrophoblastic differentiation. Our case was distinctly different from the latter because it lacked multinucleated cells and was negative for HCG.\nMin and Gillies7 reported a multinucleated giant cell stromal tumor of the omentum in an 80-year-old female patient. The immunophenotypic characteristics (strongly positive for vimentin, moderately positive for actin, and negative for cytokeratin, S100 protein, and CEA) of this tumor contrasted with those in our case. Distinction between the 2 tumors is necessary because of their apparently dissimilar biologic behavior: the patient in Min and Gillies' report remained disease free 5 years after a complete resection, but our patient died within 1 month after surgery.\nIn a review article, Young8 mentioned the term seroanaplastic carcinoma to refer to a carcinoma with easily recognizable serous differentiation admixed with areas of anaplastic morphology. Our patient's tumor could have been classified as an extremely poorly differentiated seroanaplastic carcinoma because it consisted of solid sheets of diffuse anaplastic giant cells with only minimal elements of serous differentiation.\nThe strong immunopositivity for AE1/AE3 and immunonegativity for CD45 confirmed that the tumor in our case was a carcinoma rather than a poorly differentiated lymphoma. Choriocarcinoma was ruled out because of (1) the lack of a dimorphic plexiform pattern of cytotrophoblasts and syncytiotrophoblasts and (2) a negative immunostain for HCG. Our patient's tumor was unlikely to be an anaplastic carcinoma associated with an ovarian mucinous cystic tumor,4 not only because it was present predominantly in the omentum and only superficially in the ovaries, but also because it lacked a mucinous component. Although carcinosarcoma (malignant mixed mullerian tumor) may show giant tumor cells, its characteristic histologic features of intimately admixed carcinomatous and sarcomatous elements were not seen in our patient's tumor. Chemotherapy is known to cause bizarre giant tumor cells; however, the anaplastic giant cell morphology in our case was not secondary to prior chemotherapy because of the lack of such a history in our patient.\nThe differential diagnoses also included metastatic tumors from other sites, particularly those in which anaplastic giant cell carcinomas have been reported. Melanoma was excluded because of immunonegativity for melanoma markers (HMB-45 and Melan A). The patient was found to have a completely negative clinical and diagnostic imaging workup for a primary extrapelvic carcinoma as well as the following results: immunopositivity for WT-1 and CA 125; and immunonegativity for GCDFP-15, CK20, CDX2, and TTF-1. For these reasons, the tumor was considered to represent a primary peritoneal carcinoma instead of a metastatic disease from another organ site, such as breast, gastrointestinal tract, pancreas, or lungs.\nThe pathogenesis of PPC has evoked considerable controversy. Some investigators have suggested that PPC arises primarily from the ovarian surface epithelium, with subsequent intraperitoneal spread.910 Support for this theory has been provided by studies showing similar histologic and immunohistochemical phenotypes between typical PPCs and serous ovarian carcinomas.2 However, the overwhelming predominance of peritoneal, rather than ovarian, tumor distribution is inconsistent with a sole derivation from the ovarian surface. Other investigators believe that PPC results from multifocal malignant transformation of the abdominopelvic coelomic epithelium.11 The multifocal development of PPC is supported by routine histopathologic observations that both the abdominopelvic mesothelium and the ovarian surface epithelium engender a similar histologic spectrum of tumors.12 Indeed, the abdominopelvic mesothelium and the ovarian surface epithelium share a common heritage. Both are derived from the coelomic epithelium in early embryonic life; both display similar light microscopic, electron microscopic, histochemical, and immunohistochemical features; and both carry the potential for transformation to various malignancies of similar or identical histologic types.13 Some investigators have suggested a monoclonal nature for PPC, although their conclusions were based on the examination of p53 gene mutations in only 2 cases.14 Most clonality studies have supported a multifocal pathogenesis of PPC.15 Because PPC occurs almost exclusively in women, hormonal influences on the susceptible mullerian system have been implicated in its development. However, estrogen and progesterone receptors are usually absent in PPCs, as shown in our case.\nIn the ovary, most anaplastic carcinomas are believed to be the result of dedifferentiation of an associated mucinous neoplasm. In our case, the presence of minor focus of serous carcinoma suggests that a serous carcinoma may have dedifferentiated into an anaplastic carcinoma in a manner analogous to mucinous tumor. Regardless of the histogenetic mechanism, a tumor of this type has not, to our knowledge, been reported to arise from the pelvic peritoneum.\nSurgical debulking followed by chemotherapy is the standard treatment for PPC. The reported median survival among women with typical PPC has ranged from 21 to 76 months.1 Our patient lived for less than 1 month after surgery. The strong and diffuse p53 immunoreactivity in our case was not a surprising finding, since increased nuclear accumulation of p53 correlates with unfavorable clinical outcomes in ovarian carcinoma. The aggressive behavior and highly malignant phenotype of our patient's tumor might have been related to the presence of diffuse anaplastic giant cells. Although studies of more cases with similar features would be required for substantiation, a tumor with diffuse anaplastic giant cells may represent a particularly aggressive form of primary peritoneal carcinoma.\nCopyright: College of American Pathologists 2008\nA, The tumor is composed of diffuse, solid sheets of highly malignant-appearing giant cells; nuclei resembled psammoma bodies at low magnification (hematoxylin-eosin, original magnification ×100). B, At high power, the tumor giant cells showed marked nuclear irregularity (hematoxylin-eosin, original magnification ×400). C, A focus of tumor with slitlike spaces is suggestive of serous differentiation (hematoxylin-eosin, original magnification ×100). D, The carcinoma focally involves the surfaces of the ovaries but not the parenchyma (hematoxylin-eosin, original magnification ×100). E, The tumor cells are immunopositive for AE1/AE3 (original magnification ×100). F, The tumor is also immunopositive for CA 125 (original magnification ×100)\nContributor Notes\nReprints: Cunxian Zhang, MD, PhD, Department of Pathology and Laboratory Medicine, Women & Infants Hospital of Rhode Island, Warren Alpert Medical School of Brown University, Providence, RI 02905 (czhang@wihri.org)","source_license":"CC0","license_restricted":false}