Document 6RL1vOOXVKXrYQxRKv4kVgmvR

Anatomic Pathology / Malignant Peritoneal Mesothelioma in Women Malignant Peritoneal Mesothelioma in Women A Study of 75 Cases With Emphasis on Their Morphologic Spectrum and Differential Diagnosis Patricia M. Baker, MD,1 Philip B. Clement, MD,2 and Robert H. Young, MD3 Key Words: Peritoneum; Malignant mesothelioma; Female DOI: 10.1309/2H0NVRERPP2LJDUA Abstract Seventy-five malignant mesotheliomas of the peritoneum in women were reviewed to highlight their morphologic spectrum. The patients rangedfrom 17 to 92 (mean, 47.4) years of age. The clinical presentation was usually abdominal or pelvic pain, abdominal swelling (sometimes due to ascites), or a pelvic mass. On microscopic examination, the majority of the tumors had only an epithelial morphology, but 4 were biphasic and 1 was sarcomatoid. The most common epithelial patterns were tubular and papillary (which often coexisted), but 5 tumors were purely diffuse; 2 had cells with abundant glassy eosinophilic cytoplasm (so-called deciduoid mesothelioma). The cells in the tubular and papillary patterns were generally cuboidal with scant to moderate amounts of eosinophilic cytoplasm. Nuclear atypia was usually only mild, although a minority of cases had moderate or even, occasionally, severe atypia. Many tumors hadfoci that, viewed in isolation, resembled so-called well-differentiated papillary mesothelioma, and accordingly that diagnosis should be made cautiously. Unusual features were lymphoid follicles (13 cases), striking myxoid stroma (5 cases), prominentfoamy histiocytes (5 cases), and a striking vascular proliferation (1 case). The varied morphology ofperitoneal malignant mesotheliomas may raise a broad differential diagnosis, but in most cases the resemblance to other tumors is limited. Histochemistry, immunohistochemistry, and electron microscopy may provide important aid, particularly when tissue is limited, but should be needed only occasionally. The histologic diagnosis of mesothelioma often is consid ered problematic when based only on routine light microscop ic findings, and much has been written on the undoubted value in some cases of immunohistochemical analysis in the differ entiation of mesothelioma from other tumors, especially extraovarian serous carcinomas in females.1-6 In our experi ence, however, features seen on H&E-stained sections usually allow differentiation of mesotheliomas from serous and other tumors. To highlight the morphologic features of peritoneal mesothelioma in females and emphasize the features that should enable it to be recognized and distinguished from other tumors, we reviewed the histologic features of 75 cases. The immunohistochemical features of the tumors were not ana lyzed because these features specifically were not the focus of this study and have been well described in the literature. Patient outcome also was not the focus of this study, because that topic was addressed in a recent study that included 15 of the cases in this report,7 and many of the cases in this series are recent with limited follow-up intervals. Materials and Methods Of the cases, 60 were from the consultation files of two of us (P.B.C. and R.H.Y.) and 10 were from the files of Robert E. Scully, MD. The remaining cases were from the surgical pathology files of the Massachusetts General Hospital, Boston (2 cases); University of British Columbia, Vancouver (1 case); and the United States-Canadian Mesothelioma Panel (2 cases). Fifteen cases were published in a study of diffuse malignant epithelial mesotheliomas of the peritoneum in women,7 and 1 case was reported in a study of deciduoid 724 Am J Clin Pathol 2005;123:724-737 DOI: 10.1309/2H0NVRERPP2LJDUA American Society for Clinical Pathology Anatomic Pathology / O Ariginal rticle mesotheliomas.8 Cases from a previous study of 9 cases of malignant mesothelioma with prominent ovarian involvement were not included because the histopathologic features of these tumors were described in detail in that report.9 Tumors fulfilling our strict criteria for the diagnosis of well-differenti ated papillary mesothelioma (see the "Discussion" section) were excluded. The number of H&E-stained sections avail able for review ranged from 1 to 31 (mean, 14). Mitotic counts were performed on 5 sets of 10 high-power fields (HPF), and the mitotic count in the highest set of 10 fields was recorded. Results Clinical and Operative Findings The patients were 17 to 92 years old (mean, 47.4 years). The clinical manifestations were known in 65 cases. Of these patients, 49 had abdominal or pelvic pain, abdominal swelling that in some cases was due to ascites, a pelvic mass, or com binations thereof. In 3 of these cases, a diagnosis of florid reactive or atypical mesothelial hyperplasia was made at ini tial operation, but laparotomy 2 months later (1 case), 3 years later (1 case), and at an unspecified interval for persistent symptoms (1 case) revealed malignant mesothelioma. In a fourth patient who initially had an acute bowel obstruction thought to be due to adhesions and was treated with a colosto my, a diagnosis of malignant mesothelioma was made on biopsy specimens obtained at the time of colostomy closure. In an additional case, a diagnosis of well-differentiated papil lary mesothelioma was made on a tumor that at a subsequent operation 9 months later was shown to be malignant based on its infiltrative features. The tumor was an incidental intraoperative finding in 13 patients who underwent an operation for other reasons. One patient with a history of ovarian endometrioid carcinoma treated by surgery and chemotherapy had mesothelioma diag nosed at the time of "second-look surgery," and in 1 patient initially given a diagnosis of stage III ovarian carcinoma that was treated with chemotherapy, subsequent reexploration and pathology review led to a change in diagnosis to malignant mesothelioma. In 1 case, the tumor was an autopsy finding. A history of asbestos exposure was given in 1 case, and 2 patients had received radiotherapy. At the time of laparotomy, widespread disease was iden tified in 59 cases. In many of these cases, the peritoneal sur faces were described as studded with tumor. In other cases, their appearance was less striking and characterized by terms such as granules, papillae, adhesions, or combinations there of. The remaining cases lacked the diffuse distribution of dis ease seen in the cases just noted. In 3 of these cases, disease was limited to a segment of bowel: one of them was described as diffusely fibrotic with no evidence of a dominant mass, another as being fibrotic with numerous adhesions on the serosal surface and within the mesentery and adjacent omen tum, and another as "bowel-to-bowel" adhesions forming a tumor-like mass. Disease limited to the omentum was found in 2 cases, and in an additional 2 cases, disease involving both ovaries and portions of the omentum was found. In another case, there was a 10-cm mass adjacent to the left ovary with focal disease involving 2 additional sites. In 3 cases, a single lesion was identified at the time of operation: a soft, red-tan, 4.0-cm right paratubal mass (1 case); a soft, tan-gray, 0.9-cm broad ligament mass (1 case); and a solid, tan, 2.2-cm nodule in the pelvic side wall (1 case). In the case found at autopsy, the abdominal and pelvic cavities were obliterated by tumor that encased the viscera. In 4 cases, the extent of the disease was unknown. Thirty-six patients were treated with a total abdominal hysterectomy and bilateral salpingo-oophorectomy, often with omentectomy and peritoneal biopsies that were accom panied by appendectomy (13 cases), bowel resection (3 cases), pelvic or para-aortic lymph node dissection (9 cases), liver biopsy (2 cases), or splenectomy (1 case). Three patients had a simple hysterectomy with peritoneal biopsies performed in 2 of them, and 1 patient had a total abdominal hysterectomy and unilateral salpingo-oophorectomy with biopsies. In 7 patients, bilateral salpingo-oophorectomy and multiple peritoneal biopsies were accompanied by an appen dectomy (1 case) and a resection of bowel (1 case). In 1 case, a single mass was excised; in 1, a single nodule was biopsied; and in 1, only a segment of bowel was removed. In 15 cases, only peritoneal biopsies were performed. In 8 cases, the extent of the surgery was unknown, and in 1 case, as noted, the examination was postmortem. Pathologic Findings Gross Findings Gross examination revealed that the tumor usually took the form of nodules or, less commonly, plaques, or papillary excrescences or, rarely, polyps that in 1 case were described as cauliflower-like. The neoplastic tissue usually was solid or occasionally solid and cystic and variously described as soft, friable, firm, or rubbery and white, red, yellow-tan, pink, or gray. In 1 unusual case, the tumor nodules were described as cartilaginous. In some cases, removed organs (including bowel and ovaries) were encased by tumor, and in 1 case, tumor invaded and compressed the bowel wall, constricting the bowel lumen. In the patient in whom the diagnosis was made at autopsy, the entire abdominal cavity was grossly obliterated by tumor and the viscera had to be removed en bloc. In 5 cases in which tumor microscopically involved the omentum, it was described grossly as having no distinct American Society for Clinical Pathology Am J Clin Pathol 2005;123:724-737 725 DOI: 10.1309/2H0NVRERPP2LJDUA Baker et al M P M W/ alignant eritoneal esothelioma in omen masses. In 1 case in which a large mass involved the colonic mesentery, lymph nodes were involved grossly by tumor. Microscopic Findings Seventy tumors had purely or predominantly epithelial morphologic features Umage 11, Umage 2(, Umage 31, Umage 41, (Image 51, (Image 61, (Image 71, (Image 81, (Image 91, Image 10(, and (Image 11(, 4 tumors were biphasic (epithelial and sarcomatoid) (Image 12(, and 1 tumor was purely sarco matoid (Image 13(. In 5 cases in the epithelial group, there was a transition from obvious epithelial patterns, which predomi nated, to areas with spindled morphologic features. When this transition was abrupt, it resulted in a striking picture because of the marked change in morphologic features, a low-grade (Image 1( Tubulopapillary pattern. A, A polypoid nodule composed of small tubules occupies most of the image, but a few papillae also are present (H&E). B, Typical small tubular pattern (H&E). (Image 2( Papillary pattern. The papillae are lined by a single layer of bland cuboidal to flattened cells with scant cytoplasm. Viewed in isolation, the picture is consistent with so-called well-differentiated papillary mesothelioma (H&E). See text for discussion. (Image 3( Branching slit-like tubules. The pattern might suggest a serous neoplasm, but note the lack of the cellular intraluminal papillae that are frequent in such neoplasms. The cytologic features also were blander than those of most serous carcinomas (H&E). (Image 4( Small clusters with focal tubular differentiation infiltrating in a fibrous stroma (H&E). 726 Am J Clin Pathol 2005;123:724-737 DOI: 10.1309/2H0NVRERPP2LJDUA American Society for Clinical Pathology Anatomic Pathology / O Ariginal rticle papillary pattern merging with a sarcomatoid pattern Umage 141. The architectural and cytologic spectra varied from case to case and within individual cases, but the greatest variation was seen in tumors that were purely epithelial. Three major epithelial patterns were identified: tubular, papillary, and solid. The most common was tubular, which was present at least focally in 56 cases, most often in combi nation with papillary architecture (Image 1), and was the predominant pattern in 24 cases (Image 4). The tubules gen erally were small and uniform and relatively evenly spaced (Image 4) but sometimes were dilated and in 5 cases were compressed and irregular in size and shape (Images 3 and 6). In occasional cases, slit-like spaces were striking and, to an extent, focally simulated the slit-like spaces of serous carci nomas. In 3 cases, closely packed tubules, sometimes with a cribriform pattern, formed well-circumscribed rounded Image 51 Papillae with hyalinized cores. Most of the papillae have become fused together (H&E). Image 6t Anastomosing tubules lined by cells that had eosinophilic cytoplasm and associated prominent lymphocytic infiltrate imparting a low-power appearance reminiscent of that seen in Warthin tumor (H&E). * v * h1. t:L -/ri* ; t* v H ;V ' Image 7t Insular and tubular patterns resembling those of ovarian sex cord tumors (H&E). American Society for Clinical Pathology V\ ' * - Tj Image 8t A, Diffuse growth of mesothelioma cells with obliteration of most of the underlying omental adipose tissue. Lymphocytic aggregates also are present (H&E). B, Highpower appearance of diffuse growth showing typical eosinophilic cytoplasm and relatively uniform nuclei (H&E). Am J Clin Pathol 2005;123:724-737 727 DOI: 10.1309/2H0NVRERPP2LJDUA Anatomic Pathology / O Ariginal rticle to patterns encountered in ovarian granulosa and Sertoli cell tumors (Image 7). In 5 tumors, only a solid growth pattern was identified. In 1 case with so-called deciduoid morphologic features (see next paragraph) the cells focally were arranged in broad intersecting bands (Image 9A). The neoplastic cells in the epithelial patterns generally were polygonal, cuboidal, or low columnar, with pale to eosinophilic cytoplasm that ranged from scant to moderate, to infrequently abundant, sometimes being glassy or squamoid. The polygonal cells often had sharp "squared-off' cytoplasmic borders that imparted a cobblestone appearance (Image 8B). Cells lining cysts usually were flattened and those covering Image 131 Desmoplastic malignant mesothelioma infiltrating omentum (H&E). papillae, low cuboidal (Image 2). In 3 cases, striking hob nailed cells were seen, and in 2 cases, they covered large areas of the omental surface. Cells with clear cytoplasm were pres ent focally in many cases, usually in small numbers, although in 1 tumor, they were conspicuous. In 1 case, the mesothelial cells had striking abundant foamy cytoplasm (Image 10), and in 1 case in which the cells had abundant eosinophilic cyto plasm, "apical snouts" were seen. The cells in the 2 deciduoid tumors were very large, round, oval, or polygonal with abun dant eosinophilic, glassy cytoplasm (Image 9B), with occa sional multinucleated cells in 1 case and, in the other case, pleomorphic, often multinucleated cells that vaguely resem bled intermediate trophoblasts. Variably sized cytoplasmic vacuoles were found in scattered tumor cells in many tumors, and in 3, they were prominent enough to result in a pattern resembling that of an adenomatoid tumor. In 5 cases, signet ring-like cells were focally conspicuous, and in 2 cases, some cells had a rhabdoid appearance with eccentric nuclei and prominent, dense, eosinophilic cytoplasm. The degree of nuclear atypia varied greatly among the epithelial-dominant tumors from mild (31 cases) (Image 8B), to mild to moderate (29 cases), and moderate to severe (10 cases). Nucleoli varied from tumor to tumor and within tumors from inconspicuous to prominent and eosinophilic but generally were more conspicuous in tumors with greater degrees of atypia. Nuclear grooves were striking in 1 tumor, which, combined with its growth patterns, resulted in a close resemblance to an ovarian sex cord tumor. Nuclear pseudoin clusions were seen in several tumors. Mitotic figures were rare to absent in 37 tumors (<1 mitotic figure per 10 HPF) and ranged from 1 to 5 mitotic figures per 10 HPF in 24 tumors Image 14^ Papillary pattern (top) with underlying sarcomatoid pattern (H&E). Image 15^ A, Nodular growth in case with prominent myxoid stroma (H&E). B, Scattered mesothelial cells growing mostly in a myxoid background (H&E). American Society for Clinical Pathology Am J Clin Pathol 2005;123:724-737 729 DOI: 10.1309/2H0NVRERPP2LJDUA Baker et al M P M W/ alignant eritoneal esothelioma in omen and from 6 to 15 mitotic figures per 10 HPF in 6 tumors. In 3 cases, mitotic counts were difficult to assess accurately owing to paucicellularity, including one case with abundant myxoid stroma and another with numerous apoptotic bodies. The stroma within the epithelial patterns varied from edematous to fibrous and, in some of the latter, was sclerotic or hyalinized. The stroma also was focally myxoid in 19 tumors and was conspicuously so in 5 of them, sometimes forming rounded nodules Umage 15AI (left). In one such case, the stroma of the entire tumor was composed of large pools of paucicellular, lightly basophilic, myxoid material. The cores of the papillae often were hyalinized, and in 5 cases, this was striking (Image 5). In 2 cases, hyalinized fibrous tissue formed spherical structures surrounded by epithelium, resulting in a vague resemblance to an adenoid cystic carcinoma. A mixed inflammatory infiltrate consisting predominantly of lymphocytes and plasma cells, but also neu trophils and occasional eosinophils, was present in almost all tumors. The inflammatory infiltrate generally was present in the stroma but occasionally was admixed with the tumor cells. Small collections of neutrophils formed microabscesses in several of the tumors, and in 5 cases, a striking predominance of neutrophils was seen, usually in association with areas of acellular necrotic debris. In 13 cases, a heavy infiltrate of lym phocytes and plasma cells, including lymphoid aggregates (Image 6 and Image 8A) and follicles, was present, typically within the omental fat or omental fibrous tissue surrounding areas of invading tumor. Foamy histiocytes, rarely containing hemosiderin pigment (2 cases), were present in 14 cases and were prominent in 5 of them. The foam cells most often were found in the cores of the papillae, sometimes expanding them, but also occasionally were found as single cells, small clus ters, or solid sheets in nonpapillary areas (Image 10). Necrosis was present at least focally (sometimes as single cell necrosis) in 27 tumors and was prominent in 8 of them. Psammoma bodies were present in 16 cases and were prominent in 6 of them (Image 11). They typically were seen in the papillary areas, but in 2 cases occurred in the solid areas. In one unusual tumor, a peculiar and very prominent reactive vascular proliferation was present diffusely throughout the tumor. In 3 cases, striking perivascular fibrosis was seen. An infiltrative pattern of invasion was seen in 45 cases, usually involving omentum, ovaries, bowel wall, or uterine serosa (in order of decreasing frequency). In the autopsy case, tumor destructively infiltrated the spleen and liver. In the omentum, the tumors typically dissected through the omental fat. On the ovarian surface, the invasion was often at the base of a papillary lesion as haphazardly arranged groups of tubules. In 1 case with striking myxoid stroma (Image 15A), rounded nodules involved the ovary and, viewed in isolation, could have been misconstrued as an unusual ovarian sex cord-stromal tumor. In 10 cases, the invasion had a pushing front involving the ovary (3 cases) or omentum (7 cases). In 1 case in which there was no obvious invasion in the slides available for review, lymph node metastases were present. Destructive invasion was not identified in 7 cases in which the tumor grew over the involved tissues in a plaque-like manner, including 1 case with a single focus of tumor confined to a structure resembling a mesothelial cyst. Invasion could not be assessed in 7 cases. Lymph node metastases were present in 6 cases; the intranodal tumor had a predominantly solid pattern, but focal tubulopapillary areas were seen in 1 case, and in 1 case, psammo ma bodies were identified. Lymphovascular space invasion, however, was not identified in any of the tumors. The 4 biphasic tumors (Image 12) contained an epithelial component that varied from less than 10% (2 cases) to 30% (1 case) to 50% (1 case). In the regions in which both epithelial and sarcomatoid components were present, they were associ ated intimately with each other. In 3 of the tumors, the epithe lial component was purely tubular, and in the fourth, it was solid and papillary. The degree of cytologic atypia in the epithelial component varied greatly among the 4 cases; in 1 case in which the epithelial component was minor, the epithe lial cells were very monotonous with minimal atypia and no mitotic activity. In the other 3 cases, the epithelial component showed moderate to severe cytologic atypia, including bizarre mononucleated and multinucleated tumor giant cells that in 1 case simulated anaplastic carcinoma. A brisk mitotic rate was seen in the epithelial component of 1 tumor, and in 3 cases, the mitotic activity was difficult to assess owing to the limited amount of epithelium. The cells in the sarcomatoid component of the biphasic tumors often were arranged in sheets, but fascicular and storiform patterns also were focally prominent, and 1 was prominently myxoid. The cellularity in these areas varied from highly cel lular to paucicellular. The cells were moderately to severely atypical and epithelioid to spindled with multinucleated tumor giant cells, large vesicular multilobated nuclei, occasional nuclear pseudoinclusions, and nucleoli that varied from incon spicuous to prominent. In 3 cases, the spindle cells focally merged with the epithelial component, and in 1 case, the epithelial and sarcomatous components were distinct, with a tendency of the spindle cells to focally condense around the epithelial component. The mitotic rate in the spindle cells usu ally was brisk. Extensive tumor necrosis was seen in the sar comatoid component of all tumors. All 4 biphasic tumors were highly infiltrative, with involvement of omental fat (3 cases), ovaries (2 cases), and the spleen and liver (1 case). The single sarcomatoid mesothelioma had diffuse storiform and fascicular patterns and extensively infiltrated omen tal fat (Image 13). Abundant collagen focally invested individ ual tumor cells, resulting in a desmoplastic appearance (Image 13). Necrosis, severe cytologic atypia, and brisk mitotic activ ity were found. 730 Am J Clin Pathol 2005;123:724-737 DOI: 10.1309/2H0NVRERPP2LJDUA American Society for Clinical Pathology Anatomic Pathology / O Ariginal rticle Discussion Peritoneal mesotheliomas are much less common than their pleural counterparts, and the literature on them predom inantly concerns a consideration of tumors occurring in males, who often have a history of asbestos exposure.10 The largest series of 82 cases, reported by Kannerstein and Churg10 in 1977, included only 6 females. Two series, however, focused on these tumors in women,7,11 both of which found no definite link with asbestos exposure. In addition, one of the studies7 found that a substantial number (40%) of peritoneal malignant mesotheliomas in women had an indolent behavior compared with their usually aggressive behavior in males.10 A few addi tional female cases are present in some other studies, such as one emphasizing presentation as lymph node metastasis12 and one emphasizing presentation as an ovarian mass.9 The pres ent study is the largest series describing the histopathologic features of peritoneal malignant mesotheliomas in females. The male/female ratio in patients with peritoneal malig nant mesotheliomas is 2:1, but because of the rarity of pleu ral mesotheliomas in women,13 the ratio of peritoneal to pleu ral mesotheliomas is higher in women (1:2) than in men (1:5).14-22 As previously described in the literature23-25 and as exemplified by 2 of our cases, there is an occasional associa tion with previous radiation therapy. The rarity of peritoneal mesotheliomas in women contributes to the confusion with other tumors, such as serous tumors, which are much more common. In our experience, this situation frequently leads the pathologist to not consider the diagnosis of mesothelioma and, when it is considered, a lack of experience with its histopathologic features can result in misdiagnosis. In our series, only a minority of the cases were submitted with a diagnosis of mesothelioma made or strongly favored; most were submitted with no diagnosis or considered to be neo plasms of other types. The distinction of malignant mesothe lioma from various nonmesothelial tumors, which can be done by evaluating routinely stained slides in almost all cases, is the focus of this discussion. Although other neoplasms can involve the peritoneum in a manner that is grossly indistinguishable from that of malig nant mesothelioma, the typical gross presentation of mesothelioma, as represented by most of the cases in this series, merits emphasis. Characteristically, there is limited invasion of the intra-abdominal organs, although occasional ly the latter can be replaced by tumor. Indeed, in rare cases, striking ovarian involvement can result in the clinical or intra operative appearance of a primary ovarian cancer with intraperitoneal spread.9 In a minority of cases, mesothelioma has a much less impressive clinical, intraoperative, or gross appearance. Indeed, in some cases, the mesothelioma might be an unexpected microscopic finding in a patient undergoing an operation with a clinical impression of another type of abdominal disorder. For example, in a recent study, the intra operative appearance of the tumor mimicked appendicitis, acute cholecystitis, or an incarcerated umbilical hernia.26 In these cases, the typical multifocal involvement of the peri toneum was not evident at the initial surgical exploration and the diagnosis subsequently was made on microscopic exami nation of the resected specimen. Similarly, we have seen a recent case (not included in the present study) in which a patient underwent laparotomy for voluminous ascites. No obvious tumor masses were encountered, but microscopic examination of an omentectomy specimen and peritoneal biopsy specimen revealed malignant mesothelioma. The results of the present study indicate that a tubular pattern, frequently accompanied by a papillary or solid pat tern (or all 3 in combination), and lesional cells that resemble normal or hyperplastic mesothelial cells represent the most frequent morphologic features of peritoneal mesothelioma. In the majority of cases, these features impart a characteristic appearance that makes the diagnosis straightforward. The tubules typically are small and usually uniformly arranged and are lined by cuboidal cells that in most cases have scant cytoplasm but occasionally have more abundant eosinophilic cytoplasm. The cells covering the papillae are similar to those lining tubules and usually show less stratification and fewer cellular buds than those encountered in serous carcinomas, as discussed later. Much less common than tumors that are tubular or tubulopapillary, with or without minor solid foci, are neoplasms with a uniform or almost uniform solid growth of cells that have more abundant eosinophilic cytoplasm than is usually seen in tumors with other patterns. In some of these cases, the appearance of the tumor can resemble ectopic decidua, lead ing to this pattern being descriptively referred to as "deciduoid mesothelioma."8,27-30 We had 2 tumors of this type in our series; 1 was included in a previous study of this subtype.8 Two of the well-known patterns of mesothelioma of the pleu ra are biphasic and sarcomatoid, but they are much less com mon in peritoneal mesothelioma31,32; only 4 of the former and 1 of the latter were present in our study. Occasional tumors with tubular or papillary patterns exhibit an abrupt transition to a sarcomatoid pattern, a finding that we distinguish from biphasic growth. In the latter, there is an intimate admixture of the epithelial and sarcomatoid patterns, whereas in the former, individual regions exhibit one or the other of the 2 patterns. The stroma in mesotheliomas with a solid pattern usually is inconspicuous, whereas tumors with a tubular or papillary pattern tend to have a more appreciable fibrous to hyalinized stroma, although even in these tumors, rarely is it striking. In particular, the papillae of mesotheliomas are more likely to have hyalinized cores than the papillae of serous tumors. Noteworthy are the occasional cases in which the stroma is conspicuous and strikingly myxoid,33 a finding that can result American Society for Clinical Pathology Am J Clin Pathol 2005;123:724-737 731 DOI: 10.1309/2H0NVRERPP2LJDUA Baker et al M P M W/ alignant eritoneal esothelioma in omen in peculiar arrangements of tumor cells (clusters and cords) within the myxoid matrix that might suggest esoteric diag noses such as yolk sac tumor or, should the cells appear spin dled, a mesenchymal neoplasm. Evidence of merging with more typical foci of mesothelioma, such as characteristic papillae, usually will resolve the problem. The differential diagnosis of malignant mesothelioma of the peritoneum is broad34-36 and obviously includes other peri toneal mesothelial lesions. A diagnosis of well-differentiated papillary mesothelioma was considered by the referring pathologist in 8 of our cases. These indolent tumors37,38 are small (most <2 cm) and generally solitary and show no inva sion of the underlying tissues. Microscopic examination reveals that they are characterized, according to our criteria, by a well-developed, orderly, papillary architecture in which the papillae are covered by a single layer of benign, mitotically inactive, mesothelial cells. In 5 cases in which this diagno sis was considered, multiple nodules or multiple sites of involvement were identified intraoperatively, a finding that should have suggested the possibility of a malignant process.11 In 1 of the 3 cases with a solitary lesion, the sur geon noted numerous "adhesions" that were not biopsied but later raised concern about the potential malignant nature of the process. As Goldblum and Hart11 correctly empha sized, it is crucial to be aware that individual foci within neoplasms that, based on their overall evaluation represent peritoneal malignant mesothelioma, might be indistinguish able from well-differentiated papillary mesothelioma (Image 2). Thus, it is imperative in this situation that most or all the lesions be removed or biopsied for microscopic examination. Evaluation of the morphologic features in the context of the overall gross and other histopathologic findings should, in practice, make the differential diagnosis of well-dif ferentiated papillary mesothelioma and diffuse malignant mesothelioma an infrequent problem. We should point out that our criteria for well-differentiated papillary mesothelioma are more restrictive than those of Daya and McCaughey.38 Indeed, the complexity of the morphologic features of some of their cases and the presence of multiple foci of disease would have led us to place them in the category of malignant mesothe lioma, not otherwise specified. The latter neoplasm might be indolent in a substantial minority of cases.7 Notably, Daya and McCaughey38 did not categorize their cases as benign, an important feature of their study that occasionally is overlooked. The rarity of malignant mesothelioma of the peritoneum has precluded the clear establishment of morphologic cate gories with differing outcomes. However, by using a common sense approach when signing out individual cases, one should comment not only on the amount of disease based on the tis sue available but also on its histologic features. It is logical to conclude that at least some well-differentiated malignant tumors will have more indolent behavior than tumors with more ominous morphologic features, as suggested by Kerrigan et al,7 although their findings did not reach statistical significance. On the other hand, as we hope our comments have already indicated, the use of "well-differentiated" termi nology may be overly reassuring if applied to tumors with multifocal, particularly bulky, disease. The most common mesothelial neoplasm in females is the adenomatoid tumor, but its distinction from malignant mesothelioma is rarely a problem because it typically is a grossly well-circumscribed, small tumor occurring in the myometrium, the fallopian tube, or, rarely, the ovary. Despite this, a malignant mesothelioma showing prominent vacuoliza tion of the cells raised the possibility of an adenomatoid tumor to the referring pathologist in 2 of our cases. In these cases, however, the intraoperative, gross, and other histologic find ings, including cytologic atypia, mitotic activity, and overt invasion, were all incompatible with a diagnosis of adenoma toid tumor. It should be noted, however, that although adeno matoid tumors are grossly well circumscribed, they might infiltrate adjacent tissues microscopically, a finding that should not lead to a misdiagnosis of malignant mesothelioma in the presence of the otherwise typical gross and microscop ic findings of adenomatoid tumor, including their typical loca tions. Rarely, a bona fide peritoneal adenomatoid tumor seems to arise out of a papillary mesothelioma as in 1 case recently shared with us Umage 161. We are aware of another such case (R.E. Scully, MD, oral communication, September 2003) in which the material is no longer available for review. A rare subset of adenomatoid tumors, those with striking cystic change, conceivably could be a consideration in the differ ential diagnosis of a cystic malignant mesothelioma. However, -- -.' . Image 161 Papillary pattern of mesothelioma (A, H&E) merging with adenomatoid tumor (B). B, Typical morphologic features of the adenomatoid tumor (H&E). 732 Am J Clin Pathol 2005;123:724-737 DOI: 10.1309/2H0NVRERPP2LJDUA American Society for Clinical Pathology Anatomic Pathology / O Ariginal rticle gross cysts are rare in malignant mesotheliomas, and there were no cases in our series that could be considered "cystic" malignant mesotheliomas. In our experience, the most realis tic differential diagnosis of the rare cystic adenomatoid tumors is with peritoneal inclusion cysts,39 a lesion that, although classified by others as "cystic mesothelioma" or in similar terms,40-42 is, in our opinion, a nonneoplastic lesion with an inflammatory or reactive basis that we categorize separately for reasons given in detail elsewhere.43 Distinguishing a florid reactive mesothelial proliferation from a malignant proliferation can be challenging, particularly if the amount of tissue present for examination is limited.34,36,43 Indeed, in 4 of our cases, the submitting pathologist strongly favored a reactive process. Rare cases of reactive peritoneal disease have manifested with ascites and at laparotomy have shown marked thickening of peritoneal surfaces,44,45 raising the consideration of malignant mesothelioma and enhancing the potential for pathologic misdiagnosis. Hyperplastic mesothelial cells, however, rarely form grossly visible nod ules. Normal mesothelial cells show well-defined cytoplasmic borders, small centrally located nuclei with an even chromatin pattern, and inconspicuous nucleoli. Mesothelial cells show ing mild reactive changes have nuclei that are slightly larger and have a slightly more vesicular chromatin pattern and still inconspicuous nucleoli. As the reactive features increase, the cells show a higher nuclear/cytoplasmic ratio, chromatin irregularity, prominent nucleoli, mitotic figures, and multinucleated cells. Although reactive mesothelial cells are more likely to be positive for desmin and less likely to be positive for epithelial membrane antigen and p53 than the cells of malignant mesothelioma,1 exceptions to these observations occur,46 limiting the value of immunohistochemical studies in this differential diagnosis. Unequivocal destructive invasion is diagnostic of malig nancy, but it must be evaluated cautiously. In the peritoneal cavity, invasion of omental fat or invasion of an organ must be distinguished from entrapment of mesothelial cells in reactive connective tissue or fibrin on their surfaces or between omen tal fat lobules. Atypical reactive mesothelial cells might grow in linear parallel arrays (which can be a helpful clue), small clusters, or form tubules.39 Necrosis, although suggestive of malignancy, is not diagnostic because it can be seen in benign mesothelial proliferations. The presence or absence of cyto logic atypia or mitotic activity also is not reliable because malignant mesothelial proliferations can be very monotonous, whereas exuberant reactive mesothelial proliferations, as noted, can show cytologic atypia and mitotic activity, although not abnormal mitoses. Marked nuclear atypia obviously favors malignant mesothelioma although, as noted, this feature is uncommon in these tumors. Similarly, desmoplasia in an organized pattern (such as storiform) strongly favors a diagno sis of malignant mesothelioma over that of a reactive fibrosis associated with mesothelial hyperplasia, but again, the rarity of this finding limits its usefulness. The most frequently proffered diagnosis in the cases sub mitted to us was papillary serous carcinomas of ovarian or peritoneal origin. Confusion is enhanced by the overlap in the clinical manifestations of mullerian epithelial tumors and peri toneal mesothelioma that often include elevation of the CA125 level.47 Two thirds of the patients in our study had abdom inal or pelvic pain, ascites, or a pelvic mass, alone or in com bination. The last finding, particularly in a postmenopausal woman, might bias the clinician or pathologist toward a pri mary ovarian neoplasm, a much more common tumor. The distinction of papillary serous carcinoma of ovarian origin from malignant mesothelioma, including those with striking ovarian involvement and rare primary ovarian mesotheliomas,9,11 usually can be done on the basis of histo logic features alone ITable 11. Although both tumors show a papillary architecture and might have slit-like spaces and psammoma bodies, both of the latter features are uncommon in mesothelioma. In addition, the papillae in serous carcino mas generally show more hierarchical branching, cellular stratification, and detached cell clusters, whereas the papillae in typical mesotheliomas are broader, lack budding, and have hyalinized cores more often than do serous carcinomas. Papillary serous tumors in general show a much greater degree of nuclear atypia, often with areas of anaplastic or bizarre nuclei, much higher mitotic rates, and a higher frequency of abnormal mitotic figures. If one were to grade peri toneal malignant mesotheliomas based on their cytologic features, most of them would be grade 1 or grade 2 with only a minority being grade 3.7 This is the converse of the spectrum of grades of peritoneal serous carcinoma. The differ ential diagnosis with serous carcinoma is more difficult, not surprisingly, in the occasional case of high-grade malignant ITable 1I Mesothelioma and Serous Carcinoma Mesothelioma Serous Carcinoma Cellular papillae Papillae with prominent, often hyalinized stromal cores Slit-like spaces Psammoma bodies Conspicuous eosinophilic cytoplasm Nuclear atypia Mitoses Cytokeratins 5 and 6, calretinin, thrombomodulin* Ber-EP4, B72.3, Leu-M1 (CD15)* Rare to occasional Frequent Rare to occasional Rare to occasional Typical Usually mild to moderate Occasional Positive Negative Typical Rare Typical Typical Rare Usually severe Frequent Negative Positive A panel of antibodies is recommended, including the 6 listed. A diagnosis based on the staining result of a single antibody is not reliable. American Society for Clinical Pathology Am J Clin Pathol 2005;123:724-737 733 DOI: 10.1309/2H0NVRERPP2LJDUA Baker et al M P M W/ alignant eritoneal esothelioma in omen mesothelioma, but even in these cases, typical foci of better differentiation often are present and helpful. Although rarely indicated, the potential role of immunohistochemical analysis and electron microscopy in this differential diagnosis is sum marized in the subsequent text. Another ovarian tumor that rarely might enter the differen tial diagnosis with malignant mesothelioma is clear cell carci noma, which rarely is of peritoneal origin.48 Clear cell carcino ma, like malignant mesothelioma, frequently has tubular and papillary patterns (including papillae with hyalinized cores), and, as exemplified by a few of our cases, malignant mesothe liomas can rarely contain clear cells or hobnail cells, two char acteristic cell types of clear cell carcinoma. The cells of clear cell carcinoma typically contain intracytoplasmic glycogen, often contain intracytoplasmic or intraluminal mucin, and gen erally show much greater cytologic atypia and mitotic activity than seen in malignant mesotheliomas. The latter may contain glycogen, but intracytoplasmic neutral mucin is almost always absent. Immunohistochemical staining and electron microscopy might be helpful in the rare case in which routine morphologic features and histochemical stains are inconclusive. In one of our cases, a diagnosis of an ovarian sex cord tumor initially was made owing to striking architectural simi larities with an adult granulosa cell tumor, including solid, microfollicular, and trabecular patterns. Many of the cells in this case also showed prominent longitudinal nuclear grooves, enhancing the diagnostic confusion. Other areas, however, were typical of mesothelioma, and no ovarian neoplasm was identified, facilitating the diagnosis. Malignant mixed mesodermal tumors (MMMTs) of the peritoneum are rare.49 Some of these tumors have arisen from foci of endometriosis, whereas others lacking this association might have arisen directly from the peritoneum. An MMMT might enter the differential diagnosis of biphasic mesothe lioma as it did in one of our cases. MMMTs usually show much greater nuclear atypia and mitotic activity of the epithe lial component than is present in the epithelial component of a biphasic mesothelioma, which usually resembles the epithelial component of a pure epithelial mesothelioma. Also, the epithelial components of an MMMT usually show typical mullerian features, most commonly endometrioid or serous. Heterologous elements might be present in MMMTs, a find ing that is extremely uncommon in malignant mesothe liomas.50 Rarely, the distinction of malignant mesothelioma from mullerian adenosarcoma might arise, although it was not a realistic issue in any of our cases. Should it be, adenosarcoma usually shows prominent intraglandular stro mal papillae of the sarcomatous component, which character istically forms periglandular cuffs, features not seen to an appreciable extent in mesothelioma. Two of our tumors had the features of a so-called deciduoid mesothelioma (Image 9), and a few other tumors had a less striking resemblance to decidual cells. Ectopic decidua rarely might form grossly visible nodules on the peritoneal surfaces or in the omentum51; patients with such findings are almost always pregnant. Microscopic evaluation reveals that the cells of ectopic decidua, in contrast with those of malig nant mesothelioma, have bland nuclear features and usually no mitotic activity. When the epithelial component of a biphasic mesothe lioma is minor or not represented in the biopsy specimen and the sarcomatous component is hyalinized extensively or shows a prominent storiform pattern with relatively bland cytologic features, as in one of our cases, a solitary fibrous tumor of the peritoneum might be considered.52-54 In contrast with the generally bland cytologic features of a solitary fibrous tumor, the sarcomatoid component of a biphasic mesothelioma shows focal moderate to marked cytologic atypia; mitotic activity, including abnormal forms; stromal invasion; and sometimes necrosis. Sampling obviously is cru cial in this situation, and the diagnosis of a benign fibromatous tumor of the peritoneum of any type should be made cautious ly and considered in the context of the gross features. In other words, a biopsy showing only bland morphologic features in a patient with peritoneal disease that has not been well sam pled does not lend itself to a confident interpretation. Of the 3 "classic" forms of mesothelioma, epithelial, biphasic, and sarcomatoid, the last is the rarest in the peri toneum and was represented by only 1 case in the present series. The gross distribution of disease is likely to be a help ful clue to the diagnosis of sarcomatoid mesothelioma rather than sarcoma, but it is in rare cases like this that the support ive findings of immunohistochemical analysis are particularly indicated, as they were in our case (Image 13). The tumor, also diagnosed as mesothelioma by Andrew Churg, MD, was pos itive for keratin and negative for carcinoembryonic antigen, Leu-M1 (CD15), Ber-EP4, and B72.3. Because of its rarity, sarcomatoid malignant mesothelioma should be diagnosed with great caution. For example, we have seen 1 sarcomatoid carcinoma that involved the peritoneum extensively and, after light microscopic examination, sarcomatoid mesothelioma was a realistic differential diagnosis. In that case, positivity for neutral mucin and the immunologic profile strongly favored carcinoma, indicating the usefulness of those techniques in problematic cases. However, even in that case, routine light microscopy was not to be discounted because careful scrutiny showed rare glands more consistent with the glands of adeno carcinoma than the tubules of mesothelioma. Several rare morphologic variants of mesothelioma, largely restricted to case reports, have been described but were not represented in pure form in our case material.55-58 One such previously reported peritoneal malignant mesothe lioma had a massive infiltrate of foamy histiocytes that with Sudan III stains was shown to be lipid.59 In 5 of our tumors, 734 Am J Clin Pathol 2005;123:724-737 DOI: 10.1309/2H0NVRERPP2LJDUA American Society for Clinical Pathology Anatomic Pathology / O Ariginal rticle foamy cells were conspicuous, but appropriate stains to doc ument intracellular lipid were not performed; based on the appearance of the cytoplasm with routine stains, we consider the presence of intracytoplasmic lipid within histiocytes to be the likely explanation for most of our cases. Neoplastic mesothelial cells themselves can be foamy (Image 10) owing to lipid accumulation, as has been described in a pulmonary mesothelioma.60 Numerous studies have indicated the value of histochemical and immunohistochemical analyses in distinguishing malig nant mesothelioma from adenocarcinoma. Adenocarcinomas typically contain neutral mucins that stain with the periodic acid-Schiff stain after diastase pretreatment (PASD). Mesotheliomas, on the other hand, secrete acid mucins (pre dominantly hyaluronic acid) that stain with alcian blue but not with PASD. Rare mesotheliomas, however, have been report ed to stain with PASD.61,62 For reasons stated earlier, immunohistochemical analy sis was not performed in most of the cases reported in this study. Many previous studies, however, have stressed the value of this technique, especially when an antibody panel is used in the diagnosis of malignant mesothelioma and in its differential diagnosis with adenocarcinoma. Mesotheliomas usually are negative for "epithelial" antigens. Of these, BerEP4, B72.3, Leu-M1 (CD15), MOC-31, and CA19-9 are the most useful in the differential diagnosis with primary peri toneal serous carcinoma.4 However, exceptions to these observations occur, as exemplified by one of the cases in the present study in which the submitting pathologist was uneasy about making the diag nosis of malignant mesothelioma in a tumor that was immunoreactive for B72.3, and yet one study found that 3.5% of mesotheliomas were positive for B72.3.5 Similarly, another study found that 11% of mesotheliomas were positive for BerEP4.4 The most useful markers that usually are present in epithelial malignant mesotheliomas are cytokeratins 5 and 6, calretinin, and thrombomodulin, although experience is large ly with pleural tumors.63 Ordonez4 found that 100% of peri toneal malignant mesotheliomas were immunoreactive for cytokeratins 5 and 6 (vs 24% of extraovarian serous carcino mas), 100% were positive for calretinin (vs 9% of serous car cinomas), and 74% were positive for thrombomodulin (vs 2% of serous carcinomas). Thus, no single marker is diagnostic in the separation of malignant mesothelioma from adenocarcino ma, and the results of a panel of antibodies should be inter preted in conjunction with the H&E and mucin stains. The value of electron microscopic examination of mesothe liomas also has been noted in the literature because these tumors characteristically have long, slender, nonbranching, bushy microvilli,64 but these features are not pathognomonic of mesothelioma, particularly when the mesothelioma is atypical.18 Moreover, some ovarian papillary serous carcinomas can show electron microscopic findings similar to the features of malignant mesothelioma. Malignant peritoneal mesothelioma in females has fre quently encountered typical patterns, which, in combination with the cell type and gross characteristics, should enable the diagnosis to be made on H&E-stained slides in most cases. As with most categories of neoplasia, aberrant morphologic fea tures occasionally are seen, and, in such cases, histochemical, immunohistochemical, and electron microscopic studies might be important aids in diagnosis and certainly are recom mended in problematic cases. The frequency of their use will, to a significant degree, be related to the confidence of individ ual pathologists with the morphologic features of mesothe lioma. Even in these cases, however, a time-honored principle of surgical pathology--extensive sampling--might have as great or greater importance by disclosing minor foci of typical mesothelioma. The clinical findings and gross characteristics of disease also should be given appropriate consideration. From the department ofPathology, Health Sciences Centre and the University ofManitoba, Winnipeg, Canada; 2Department of Pathology, Vancouver General Hospital and the University of British Columbia, Vancouver, Canada; and 3James Homer Wright Pathology Laboratories of the Massachusetts General Hospital and Department ofPathology, Harvard Medical School, Boston, MA. Address reprint requests to Dr Baker: Dept ofPathology, Health Sciences Centre, 820 Sherbrook St, Winnipeg, Manitoba R3A 1R9 Canada. References 1. Attanoos RL, Webb R, Dojcinov SD, et al. Value of mesothelial and epithelial antibodies in distinguishing diffuse peritoneal mesothelioma in females from serous papillary carcinoma of the ovary and peritoneum. Histopathology. 2002;40:237-244. 2. Bollinger DJ, Wick MR, Dehner LP, et al. Peritoneal malignant mesothelioma versus serous papillary adenocarcinoma: a histochemical and immunohistochemical comparison. Am J Surg Pathol. 1989;13:659-670. 3. Khoury N, Raju U, Crissman JD, et al. A comparative immunohistochemical study of peritoneal and ovarian serous tumors, and mesotheliomas. Hum Pathol. 1990;21:811-819. 4. Ordonez NG. Role of immunohistochemistry in distinguishing epithelial peritoneal mesotheliomas from peritoneal and ovarian serous carcinomas. Am J Surg Pathol. 1998;22:12031214. 5. Riera JR, Astengo-Osuna C, Longmate JA, et al. The immunohistochemical diagnostic panel for epithelial mesothelioma: a reevaluation after heat-induced epitope retrieval. Am J Surg Pathol. 1997;21:1409-1419. 6. Wick MR. immunophenotyping of malignant mesothelioma. Am J Surg Pathol. 1997;21:1395-1398. 7. Kerrigan SA, Turnnir RT, Clement PB, et al. Diffuse malignant epithelial mesotheliomas of the peritoneum in women: a clinicopathologic study of 25 patients. Cancer. 2002;94:378-385. American Society for Clinical Pathology Am J Clin Pathol 2005;123:724-737 735 DOI: 10.1309/2H0NVRERPP2LJDUA Baker et al M P M W/ alignant eritoneal esothelioma in omen 8. Nascimento AG, Keeney GL, Fletcher CD. Deciduoid peritoneal mesothelioma: an unusual phenotype affecting young females. Am J Surg Pathol. 1994;18:439-445. 9. Clement PB, Young RH, Scully RE. Malignant mesotheliomas presenting as ovarian masses: a report of nine cases, including two primary ovarian mesotheliomas. Am J Surg Pathol. 1996;20:1067-1080. 10. Kannerstein M, Churg J. Peritoneal mesothelioma. Hum Pathol. 1977;8:83-94. 11. Goldblum J, Hart WR. Localized and diffuse mesotheliomas of the genital tract and peritoneum in women; a clinicopathologic study of nineteen true mesothelial neoplasms, other than adenomatoid tumors, multicystic mesotheliomas, and localized fibrous tumors. Am J Surg Pathol. 1995;19:1124-1137. 12. Sussman J, Rosai J. Lymph node metastasis as the initial manifestation of malignant mesothelioma: report of six cases. Am J Surg Pathol. 1990;14:819-828. 13. Legha SS, Muggia FM. Pleural mesothelioma: clinical features and therapeutic implications. Ann Intern Med. 1977;87:613-621. 14. Asensio JA, Goldblatt P, Thomford NR. Primary malignant peritoneal mesothelioma: a report of seven cases and a review of the literature. Arch Surg. 1990;125:1477-1481. 15. Fox H. Primary neoplasia of the female peritoneum. Histopathology. 1993;23:103-110. 16. Jones DE, Silver D. Peritoneal mesotheliomas. Surgery. 1979;86:556-560. 17. Kannerstein M, Churg J, McCaughey WT, et al. Papillary tumors of the peritoneum in women: mesothelioma or papillary carcinoma. Am J Obstet Gynecol. 1977;127:306-314. 18. McCaughey WT, Colby TV, Battifora H, et al. Diagnosis of diffuse malignant mesothelioma: experience of a US/Canadian Mesothelioma Panel. Mod Pathol. 1991;4:342-353. 19. Piccigallo E, Jeffers LJ, Reddy KR, et al. Malignant peritoneal mesothelioma: a clinical and laparoscopic study of ten cases. DigDis Sci. 1988;33:633-639. 20. Plaus WJ. Peritoneal mesothelioma. Arch Surg. 1988;123:763-766. 21. Roggli VL, Oury TD, Moffatt EJ. Malignant mesothelioma in women. In: Fechner RE, Rosen PP, eds. Anatomic Pathology. Chicago, IL: ASCP Press; 1997:147-163. 22. Swan N, Cottell DC, Sheahan K. Peritoneal mesotheliomas. Am J Surg Pathol. 1997;21:122-124. 23. Beier KM, Gallup DG, Burgess R, et al. Occurrence of malignant peritoneal mesothelioma after surgery and irradiation for cervical cancer. Gynecol Oncol. 1984;17:375-380. 24. Gilks B, Hegedus C, Freeman H, et al. Malignant peritoneal mesothelioma after remote abdominal radiation. Cancer. 1988;61:2019-2021. 25. Weissmann LB, Corson JM, Neugut AI, et al. Malignant mesothelioma following treatment for Hodgkin's disease. J Clin Oncol. 1996;14:2098-2100. 26. Kerrigan SA, Cagle P, Churg A. Malignant mesothelioma of the peritoneum presenting as an inflammatory lesion: a report of four cases. Am J Surg Pathol. 2003;27:248-253. 27. Ordonez NG. Epithelial mesothelioma with deciduoid features: report of four cases. Am J Surg Pathol. 2000;24:816-823. 28. Shanks JH, Harris M, Banerjee SS, et al. Mesotheliomas with deciduoid morphology: a morphologic spectrum and a variant not confined to young females. AmJ Surg Pathol. 2000;24:285-294. 29. Shia J, Erlandson RA, Klimstra DS. Deciduoid mesothelioma: a report of 5 cases and literature review. Ultrastruct Pathol. 2002;26:355-363. 30. Talerman A. Deciduoid or pseudodecidual mesothelioma [letter]. Am J Surg Pathol. 2000;24:1179. 31. Cantin R, Al-Jabi M, McCaughey WT. Desmoplastic diffuse mesothelioma. Am J Surg Pathol. 1982;6:215-222. 32. Colby TV. The diagnosis of desmoplastic malignant mesothelioma [editorial]. Am J Clin Pathol. 1998;110:135-136. 33. Diaz LK, Okonkwo A, Solans EP, et al. Extensive myxoid change in well-differentiated papillary mesothelioma of the pelvic peritoneum. Ann Diagn Pathol. 2002;6:164-167. 34. Churg A, Colby TV, Cagle P, et al. The separation of benign and malignant mesothelial proliferations. Am J Surg Pathol. 2000;24:1183-1200. 35. Clement PB. Reactive tumor-like lesions of the peritoneum. Am J Clin Pathol. 1995;103:673-676. 36. Wick MR, Mills SE. Mesothelial proliferations: an increasing morphologic spectrum. Am J Clin Pathol. 2000;113:619-622. 37. Butnor KJ, Sporn TA, Hammar SP, et al. Well-differentiated papillary mesothelioma. Am J Surg Pathol. 2001;25:1304-1309. 38. Daya D, McCaughey WT. Well-differentiated papillary mesothelioma of the peritoneum: a clinicopathologic study of 22 cases. Cancer. 1990;65:292-296. 39. McFadden DE, Clement PB. Peritoneal inclusion cysts with mural mesothelial proliferation: a clinicopathological analysis of six cases. Am J Surg Pathol. 1986;10:844-854. 40. Alvarez-Fernandez E, Rabano A, Barros-Malvar JL, et al. Multicystic peritoneal mesothelioma: a case report. Histopathology. 1989;14:199-208. 41. Weiss SW, Tavassoli FA. Multicystic mesothelioma: an analysis of pathologic findings and biologic behavior in 37 cases. Am J Surg Pathol. 1988;12:737-746. 42. Ross MJ, Welch WR, Scully RE. Multilocular peritoneal inclusion cysts (so-called cystic mesotheliomas). Cancer. 1989;64:1336-1346. 43. Sugarbaker PH, Yan H, Grazi RV, et al. Early localized peritoneal mesothelioma as an incidental finding at laparoscopy: report of a case and implications regarding natural history of the disease. Cancer. 2000;89:1279-1284. 44. Finney AL, Spagnolo DV, Crawford GP, et al. Pseudosarcomatous sclerosing peritonitis: a case report on an unusual form of chronic lupus peritonitis. Int J Surg Pathol. 1996;4:121-128. 45. Riddell RH, Goodman MJ, Moossa AR. Peritoneal malignant mesothelioma in a patient with recurrent peritonitis. Cancer. 1981;48:134-139. 46. Kung IT, Thallas V, Spencer EJ, et al. Expression of muscle actins in diffuse mesotheliomas. Hum Pathol. 1995;26:565-570. 47. Almudevar Bercero E, Garcia-Rostan Perez GM, Garcia Bragado F, et al. Prognostic value of high serum levels of CA125 in malignant secretory peritoneal mesotheliomas affecting young women: a case report with differential diagnosis and review of the literature. Histopathology. 1997;31:267-273. 48. Lee KR, Verma U, Belinson J. Primary clear cell carcinoma of the peritoneum. Gynecol Oncol. 1991;41:259-262. 49. Choong SY, Scurry JP, Planner RS, et al. Extrauterine malignant mixed mullerian tumor of primary peritoneal origin. Pathology. 1994;26:497-498. 50. Chave G, Chalabreysse L, Picaud G, et al. Malignant pleural mesothelioma with osteoblastic heterologous elements: CT and MR imaging findings. AJR Am J Roentgenol. 2002;178:949-951. 51. Clement PB, Young RH, Scully RE. Nontrophoblastic pathology of the female genital tract and peritoneum associated with pregnancy. Semin Diagn Pathol. 1989;6:372-406. 736 Am J Clin Pathol 2005;123:724-737 DOI: 10.1309/2H0NVRERPP2LJDUA American Society for Clinical Pathology Anatomic Pathology / O Ariginal rticle 52. Fukunaga M, Naganuma H, Nikaido T, et al. Extrapleural solitary fibrous tumor: a report of seven cases. Mod Pathol. 1997;10:443-450. 53. Hill RP. Malignant fibrous mesothelioma of the peritoneum. Cancer. 1953;6:1182-1185. 54. Young RH, Clement PB, McCaughey WTE. Solitary fibrous tumors ("fibrous mesotheliomas") of the peritoneum: a report of three cases and a review of the literature. Arch Pathol Lab Med. 1990;114:493-495. 55. Henderson DW, Attwood HD, Constance TJ, et al. Lymphohistiocytoid mesothelioma: a rare lymphomatoid variant of predominantly sarcomatoid mesothelioma. Ultrastruct Pathol. 1988;12:367-384. 56. Khalidi HS, Medeiros LJ, Battifora H. Lymphohistiocytoid mesothelioma: an often misdiagnosed variant of sarcomatoid malignant mesothelioma. Am J Clin Pathol. 2000;113:649-654. 57. Kitazawa M, Kaneko H, Toshima M, et al. Malignant peritoneal mesothelioma with massive foamy cells: Codfish roe-like mesothelioma. Acta Pathol Jpn. 1984;34:687-692. 58. Ordonez NG, Mackay B. Glycogen-rich mesothelioma. Ultrastruct Pathol. 1999;23:401-406. 59. Chang HT, Yantiss RK, Nielsen GP, et al. Lipid-rich diffuse malignant mesothelioma: a case report. Hum Pathol. 2000;31:876-879. 60. Cavazza A, Pasquinelli G, Agostini L, et al. Foamy cell mesothelioma. Histopathology. 2002;41:369-371. 61. Cook DS, Attanoos RL, Jalloh SS, et al. "Mucin-positive" epithelial mesothelioma of the peritoneum: an unusual diagnostic pitfall. Histopathology. 2000;37:33-36. 62. MacDougall DB, Wang SE, Zidar BL. Mucin-positive epithelial mesothelioma. Arch Pathol Lab Med. 1992;116:874880. 63. Miettinen M, Sarlomo-Rikala M. Expression of calretinin, thrombomodulin, keratin 5, and mesothelin in lung carcinomas of different types: an immunohistochemical analysis of 596 tumors in comparison with epithelioid mesotheliomas of the pleura. Am J Surg Pathol. 2003;27:150158. 64. Dardick I, Al-Jabi M, McCaughey WT, et al. Ultrastructure of poorly differentiated diffuse epithelial mesotheliomas. Ultrastruct Pathol. 1984;7:151-160. American Society for Clinical Pathology Am J Clin Pathol 2005;123:724-737 737 DOI: 10.1309/2H0NVRERPP2LJDUA