Endometrioid Adenocarcinoma with Prominent Component of Spindled Cells Robert H. Young Massachusetts General Hospital Boston, MA

Clinical Summary: A 61-year -old woman with abdominal pain and swelling was found to have a left adnexal mass. Laparotomy disclosed a left ovarian tumor which was removed along with the contralateral adnexa.Staging was negative. The left ovary was replaced by a 13cm in greatest dimension mass which on sectioning was composed of soft, solid tissue with abundant hemorrhage. Slide 1 Click to view with ImageScope Click to view with a Web-Based Viewer Figure 1 Striking spindle cell proliferation. Figure 2 Undulating growth of neoplastic cells, most of which have a spindle cell morphology. Figure 3 High-power view of prior illustrations showing focal vague tendency to form whorls. Figure 4 Metaplastic bone. Figure 5 Densely cellular aggregates of spindle cells separated by a looser myxoid component. Figure 6 Myxoid component. Figure 7 Lower-power view showing typical endometrioid glands in association with cellular foci similar to compact regions of previous illustrations. Figure 8 Typical endometrioid glands. Figure 9 Irregularly shaped epithelial formations. Figure 10 Spindle cells merging with foci of abortive gland differentiation. Figure 11 Well-delineated nests within background fibrous stroma. Figure 12 Low-power showing epithelial arrangements as well as spindle cell component with myxoid background. Diagnosis: Endometrioid Adenocarcinoma with Prominent Component of Spindled Cells. Discussion: I will use this case as a focal point to discuss some interesting aspects of the evaluation of ovarian tumors with a predominant, to exclusive, component of spindle cells. There are many potential issues as the two tables will indicate but only a few of particular note can be considered in the time/space available. The case itself is a striking example of a form of endometrioid carcinoma in which there is a remarkable component of spindled cells, almost certainly spindled epithelial cells, and about which the M.D. Anderson Group and our group wrote a paper some years ago now [1]. We wrote the paper because of the importance of distinguishing this lesion from the malignant mixed mesodermal tumor (carcinosarcoma) of the ovary. As may come out to varying degrees from some of the other presentations, according to the focus of the speaker's presentation, the appreciation of a prominent spindle cell morphology connotes quite different things in one site or organ compared to another. Certainly in the ovary when I see a prominent spindle cell pattern, particularly when it is clearly not a pure pattern, the diagnosis of an endometrioid carcinoma quickly becomes a possibility and a judicious search is obviously undertaken to find definitive foci of endometrioid neoplasia as are clearly present in this case, albeit overwhelmed in most regions by the spindle cell morphology. When Dr. Robert E. Scully and I presented a consideration of the differential diagnosis of ovarian tumors based on their patterns and cell types in 2001 [2], spindle cells were the second cell type considered (after the ubiquitous small cell tumors), because in the ovary a spindle cell picture presents not only a wide range of considerations but also a very interesting one. The reader is referred to the paper just mentioned and an excellent review of the topic of spindle cell lesions of the ovary by Irving and McCluggage [3] for more detailed remarks. A prominent spindle cell component in endometrioid carcinoma often co-exists with another facet of endometrioid carcinoma that can be problematic, namely, sex cord-like epithelial formations. The spindle cells on that background can be readily considered the stromal component of a sex cord-stromal tumor. The spindle cells in endometrioid carcinomas can be quite varied in their appearance. Sweeping fascicles as seen in various primary mesenchymal neoplasms are common but sometimes the cells are more separated and there is a looser appearance with even, in some cases, a myxoid background. In other cases they form rounded whorls in which spindled to fusiform cells almost certainly reflecting abortive squamous cells are a major clue to the diagnosis of an endometrioid carcinoma. Foci of more overt squamous differentiation may be present if carefully searched for. Even more important than the differential diagnosis with a sex cord-stromal tumor is the differential with a malignant mixed mesodermal tumor (MMMT). The latter is of course a tumor of epithelial cell lineage but is a much more highly malignant tumor than the endometrioid carcinoma with a spindle cell population that is almost certainly not mesenchymal, and of course the cells have to be to unequivocally sarcomatous to make the diagnosis of MMMT. Although there may be cases where the distinction gets rather blurred because in my opinion there is an imprecise line between high-grade endometrioid carcinoma with spindled epithelial cells and the MMMT, in the great majority of cases one can make the distinction on H&E evaluation. Most of the endometrioid carcinomas with spindled epithelial cells are grade 2 of 3 carcinomas which have a more orderly arrangement of the various components than seen in the MMMT. This is all discussed in detail in our original paper on the topic. Although one may perform immunohistochemistry for cytokeratin, candidly it is usually not definitive in my opinion although certainly strong labeling of spindle cells for cytokeratin is in favor of spindle cell carcinoma as opposed to a true sarcomatous component. Further complicating the evaluation of spindle cells in an endometrioid neoplasm is the fact that sometimes the background fibromatous stroma of such neoplasms which, albeit part of the tumor, is not inherently neoplastic, is cellular and even mitotically active. For example, some tumors which are otherwise simple endometrioid adenofibromas can have a cellular stroma with some mitotic activity which occasionally leads to an erroneous diagnosis of the low-grade form of MMMT, the mesodermal adenosarcoma. Lack of the typical architectural features of the adenosarcoma and exclusively bland nuclear features are clues that one is just looking at an adenofibroma with an unusually cellular stroma. As I have elected an endometrioid neoplasm for my case it is appropriate to briefly mention some of the issues that come up with another spindle cell endometrioid neoplasm, endometrial stromal sarcoma of the uterus metastatic in the ovary or the even less common primary endometrial stromal sarcoma of the ovary, arising most of the time out of the background of endometriosis. Apart from the history (helpful in cases of metastasis) or associated endometriosis (helpful in the primary cases), one gets clues from a variety of features in addition to the frequent bilaterality of the metastatic tumors. Although the neoplastic cells are spindled they are not uniformly as elongated as in many other spindle cell neoplasms and range from spindled to fusiform. Unfortunately when one sees this tumor in the ovary the classic tongue-like growth seen in the uterus and often in extra-ovarian sites, is rarely conspicuous and numerous issues in differential diagnosis can come up. We have considered these elsewhere in papers on endometrioid stromal sarcomas of the ovary [4] overall and in our paper on sarcomas metastatic to the ovary [5], the greatest number of which were endometrial stromal sarcomas. An appreciation that an intensely cellular sarcomatous neoplasm has rather bland cytology can be an initial clue to the diagnosis and when that is buttressed by finding the typical prominent vascularity one can very quickly arrive at the diagnosis, or at least seriously entertain, endometrioid-endometrial stromal neoplasia. It must be noted that the vascularity is often not as conspicuous as one might hope it would be. The differential diagnosis with a sex cord tumor such as a granulosa cell tumor with a diffuse pattern may be aided by negative inhibin and calretinin staining as long as one is aware that some classic granulosa cell tumors are negative for inhibin and of course the issue gets muddied by the fact that endometrial stromal neoplasms may show sex cord-like differentiation which may show reactivity for inhibin. To say diagnostic problems here may be a challenge is an understatement! Strong staining for CD10 is supportive of an endometrioid stromal sarcoma although we have found CD10 a somewhat nonspecific stain. Prominent staining for estrogen and progesterone receptor is in favor of an endometrial stromal neoplasm. The only other primary ovarian tumor with a spindle cell morphology I will briefly note is the mucinous tumor of the ovary. Occasionally the mural nodules in such tumors, which typically are composed of large epithelial cells with abundant eosinophilic cytoplasm, have a striking spindle cell pattern [6]. Finally, I will conclude my consideration by briefly noting that of the small group of metastatic tumors that may have a striking spindle cell morphology, one to be on the lookout for, and indeed it has tricked us, is metastatic gastrointestinal stromal tumor [7]. There can of course be important therapeutic implications to its recognition. Table 1 - Primary Ovarian Tumors With Prominent Spindle Cells Fibromatous tumors (typical, cellular, mitotically active, and fibrosarcomas) Sex cord-stromal tumors Malignant mixed mesodermal tumors Endometrioid carcinoma Endometrioid stromal sarcoma Sarcomatoid foci in mucinous tumors Myxoma Miscellaneous others Table 2 - Metastatic Ovarian Tumors With Prominent Spindle Cells Endometrial stromal sarcoma Gastrointestinal stromal tumors Malignant melanoma (infrequently) References Tornos C, Silva EG, Ordonez NG, Gershenson DM, Young RH, Scully RE. Endometrioid carcinoma of the ovary with a prominent spindle-cell component, a source of diagnostic confusion. A report of 14 cases. Am J Surg Pathol 19: 1943-1353, 1995. Young, RH, Scully RE. Differential diagnosis of the ovarian tumors based primarily on their patterns and cell types. Semin Diagn Pathol 18, 161-235, 2001. Irving JA, McCluggage WG. Ovarian spindle cell lesions. A review with emphasis on recent developments and differential diagnosis. Adv Anat Pathol 14:305-319, 2007. Young RH, Prat J, Scully RE. Endometrioid stromal sarcomas of the ovary. A clinicopathologic analysis of 23 cases. Cancer 53:1143-1155, 1984. Young RH, Scully RE. Sarcomas metastatic to the ovary: A report of 21 cases. Int J Gynecol Pathol 9:231-252, 1990. Provenza C, Young RH, Prat J. Anaplastic carcinoma in mucinous ovarian tumors. A clinicopathologic study of 34 cases emphasizing the crucial impact of stage on prognosis, their histologic spectrum, and overlap with sarcomalike mural nodules. Am J Surg Pathol 32:383-389, 2008. Irving JA, Lerwill MF, Young RH. Gastrointestinal stromal tumors metastatic to the ovary. A report of five cases. Am J Surg Pathol 29:920-926, 2005.