—  SPECIALTY CONFERENCE  —

Gynecologic Pathology

Case 2 - Primitive Neuroectodermal Tumor of the Uterus

Esther Oliva
Massachusetts General Hospital


Click on each slide thumbnail image for an enlarged view
Clinical history:
A 26-year-old female with no relevant past medical history presented with uterine bleeding of 3 weeks duration. An endometrial curettage was interpreted as malignant neoplasm most consistent with sarcoma. The patient underwent radical hysterectomy with extensive lymph node sampling.


Case 2 - Slide 1
Click to view with ImageScope
Click to view with a Web-Based Viewer



Case 2 - Figure 1A
At low power magnification, the infiltrating border of the tumor is lobulated in some areas (A) while has an irregular destructive growth in other areas (B).
Case 2 - Figure 1B
At low power magnification, the infiltrating border of the tumor is lobulated in some areas (A) while has an irregular destructive growth in other areas (B).
Case 2 - Figure 2A
At higher magnification, the tumor is hypercellular showing a diffuse growth of small "blue" cells which focally seem to arrange around acellular pink areas

Case 2 - Figure 2B
At higher magnification, the tumor is hypercellular showing a diffuse growth of small "blue" cells which focally seem to arrange around acellular pink areas
Case 2 - Figure 3A
The vasculature associated with the tumor is not striking with scattered medium-sized vessels present throughout.
Case 2 - Figure 3B
The tumor cells do not show prominent degree of pleomorphism.

Pathologic findings:
The tumor measured 5 x 4 x 3 cm. Sectioning revealed a heterogeneous surface with fleshy white to gray areas alternating with areas of necrosis. On microscopic examination, the tumor was composed of small blue cells infiltrating the myometrium with a lobulated growth. At higher magnification, the small cells were tightly packed and grew in sheets. They had scant cytoplasm, oval hyperchromatic nuclei with inconspicuous nucleoli and variable mitotic activity. Only focally, very poorly formed pseudorosettes with eosinophilic central material were observed. The tumor cells were positive for vimentin, keratin, CD99 and FLI-1 and negative for chromogranin, CD10, desmin, h-caldesmon and myogenin. FISH analysis demonstrated the EWS/FLI-1 chimeric transcript, t(11;22)(q24;q12).

Diagnosis:
PNET of the uterus

Primitive neuroectodermal tumor (PNET) is a rare tumor originated from fetal neuroectodermal cells. Although originally described by Stout in 1918, the term PNET was coined by Hart and Earle in 1973 to describe a group of small round cell tumors possibly derived from fetal neuroectodermal cells showing varying signs of neural, glial, and ependymal differentiation. These tumors occur typically in the central nervous system (involving brain and spinal cord) but also may occur in peripheral locations (pPNET) often involving peripheral nerves or other sites. Peripheral PNET are currently accepted as the most differentiated form of tumor with in the PNET/Ewing sarcoma family. Similar to their counterpart in the central nervous system there are three main morphologic subtypes: differentiated (most commonly ependymoma), primitive, and anaplastic (resembling glioblastoma multiforme). In the female genital tract, the most common location is the ovary (more commonly arising from immature teratomas). Pure PNET in the female genital tract are extremely rare with less than 20 cases published in the uterine corpus and less than 5 in the uterine cervix while they are very rare in the vulva/vagina.

Clinical features
Uterine PNETs were first described by Hendrickson and Scheithauer in 1986. They can occur in a wide age range (12 to 78 years) but > 75% of the patients are postmenopausal. Patients typically present with abnormal vaginal bleeding and have an enlarged uterus or palpable pelvic mass and not infrequently at the time of initial diagnosis the tumor is already at an advanced stage.

Pathologic features
Grossly, the tumors appear as unencapsulated and bulky with a gray-tan fleshy cut surface and areas of hemorrhage and necrosis are frequently encountered. On low power microscopic examination, the tumors are frequently multilobulated and densely cellular separated by broad septa. They are composed of sheets or nests of uniform small "blue" cells with little intervening stroma. Abortive rosette-like structures or Homer-Wright pseudorosettes may be focally identified. The tumor cells may also form microcysts and follicle-like spaces. The cells have scant cytoplasm, small nuclei with evenly distributed delicate chromatin, inconspicuous nucleoli and variable mitotic activity. Variable degrees of neural, glial, ependymal and medulloepithelial differentiation may be seen, although in most cases it has been minimal. Rarely, uterine PNETs have been reported in association with endometrioid carcinoma (2 cases) and endometrial stromal sarcoma (1 case) and one PNET has been reported to show cartilaginous differentiation (histologically benign appearing).

Ancillary studies
From the histochemical and immunohistochemical point of view these tumors are PAS positive as they contain glycogen; they also stain for Vimentin, CD99 (also known as MIC2, it is a very sensitive but not very specific marker for PNET which typically shows membranous staining) and NSE. The tumor cells show variable positivity for Synaptophysin, Chromogranin, Leu-7, CD57, GFAP and S-100. FLI-1, part of the fusion protein EWS-FLI1 is present in more than 90% of tumors diagnosed as PNET (nuclear staining) and it is more specific than CD99 in the diagnosis of pPNET. A variable percentage of these tumors also express CD117, although this finding has uncertain significance.

Electron-microscopy reveals primitive mesenchymal cells with rudimentary junctions and paucity of cytoplasmic organelles. The cells may contain aggregates of glycogen but usually not neurosecretory granules.

These tumors show the EWS/FLI-1 chimeric transcript, derived from the t(11;22)(q24;q12) chromosomal translocation. The translocation can be detected by both reverse transcriptase-polymerase chain reaction (PCR) and fluorescence in situ hybridization (FISH). It has been shown that FISH has a higher sensitivity (91%) and specificity (100%) than PCR. However, it is important to remember that this translocation is not diagnostic of PNET as it has been demonstrated in other neoplasms.

Differential Diagnosis
Uterine PNETs have been reported to be frequently misdiagnosed as poorly differentiated sarcomas or carcinomas. Thus, these tumors should be distinguished from other tumors composed of small cells including low-grade endometrial stromal sarcoma, leiomyosarcoma (less likely with a highly cellular leiomyoma), rhabdomyosarcoma, small cell carcinoma, lymphoma, and malignant mixed mullerian tumor with extensive neuroectodermal differentiation. Other entities that may enter into consideration when the tumor is more differentiated include pure glioma, teratoma with glial tissue, isolated glial tissue and rarely metastases.

A- Low-grade endometrial stromal sarcoma may enter in the differential diagnosis at low power as: a) the tumor infiltrates the myometrium with tongues or lobules that simulate the infiltration pattern that may be seen in a PNET; b) it is hypercellular and uniform; and c) it is composed of small blue cells, features that may overlap with those seen in PNETs. However, at higher magnification, the tumor cells are less primitive and mitotic activity tends to be low. Characteristic arterioles reminiscent of those seen in proliferative endometrium are also seen and the tumor cells may be seen whorling around them. The tumor may also contain histiocytes and collagen bands. Finally the cut surface of these tumors is homogeneously tan to yellow in contrast to the fleshy appearance of PNETs. Immunohistochemical stains are helpful as endometrial stromal tumors are typically positive for CD10 and estrogen and progesterone receptors. Both tumors may be positive for keratin, and although sex cord areas in endometrial stromal tumors have been reported positive for CD99, pure endometrial stromal tumors are negative for this marker while PNET are typically positive. Endometrial stromal tumors are also negative for FLI-1 and neuroendocrine markers.

B- Smooth muscle tumors in the differential diagnosis include a highly cellular leiomyoma and leiomyosarcoma. Highly cellular leiomyoma enters in this differential diagnosis only at low-power examination because of the "blue" appearance of the tumor; however, all other characteristic features of this benign smooth muscle tumor are missing including bland elongated nuclei, scant to absent mitotic activity, large thick blood vessels, cleft-like spaces, fascicular growth (at least present focally more commonly at the periphery of the tumor) as well as focal merging with the surrounding myometrium. The gross appearance is also quite different from PNET as he former typically shows a tan to yellow cut surface in contrast to the fleshy cut surface seen in PNET. Leiomyosarcomas with small cells and without a fascicular growth are rare. However, in small samples the differential diagnosis may be difficult. Leiomyosarcomas typically show a fascicular growth at least focally and the cells have elongated nuclei with variable degrees of pleomorphism and variable amounts of eosinophilic cytoplasm. Weak CD99 positivity has been rarely reported in epithelioid smooth muscle tumors, PNET may express focally actin, and keratins may be seen in both tumor types; however, desmin (more sensitive) and h-caldesmon (more specific) are typically extensively positive in smooth muscle tumors while only one PNET of the soft tissues has been reported to show focal desmin positivity. If positive, estrogen and progesterone receptors may also be also helpful in this distinction (in smooth muscle tumors) while FLI-1 positivity has not been reported in smooth muscle tumors of the uterus.

C- Primary rhabdomyosarcoma of the uterus is rare with the pleomorphic variant being more common in the uterine corpus and the embryonal rhabdomyosarcoma predominating in the cervix (sarcoma botryoides). The former is easy to distinguish from PNET as the cells are large, pleomorphic and show abundant eosinophilic cytoplasm. The differential diagnosis with embryonal rhabdomyosarcoma should not create problems in a large specimen, as sarcoma botryoides typically shows a cambium layer under the surface epithelium and around the glands; it typically shows hyper and hypocellular areas and rhabdomyoblasts with cross striations may be found although they are not required to establish this diagnosis. The differential diagnosis is more difficult in biopsy specimens due to the limitation of the sample which may not show the characteristic features of this tumor. In these cases helpful immunohistochemical stains include myoglobin, MyoD-1 and myogenin. It is important to remember that CD99 is not a pathognomonic marker of PNET as it can also be positive in rhabdomyosarcoma, lymphoma and sex cord areas of endometrial stromal tumors as mentioned earlier.

D- Small cell carcinoma is a small blue tumor that also grows in sheets and may form pseudo-rosettes. However, the tumor cells often show molding and prominent apoptosis. The nuclei have the classic stippled "salt and pepper" chromatin and not infrequently one can see "crushed artifact". More frequently than PNET, small cell carcinoma is seen in association with another endometrial carcinoma component. Immunohistochemistry for neuroendocrine markers shows overlap between the two tumors. Diffuse positivity for keratin favors a diagnosis of small cell carcinoma while CD99 staining supports the diagnosis of PNET, although it is important to keep in mind that rarely small cell carcinomas may be CD99 positive.

E- Lymphoma involving the uterus should enter in the differential diagnosis as it may be composed of small blue cells. However, the tumor tends to grow in sheets with no apparent lobulation and at high power magnification the cells have a more variable appearance, may be associated with crushed artifact and prominent sclerosis. Immunohistochemical stains may be helpful in limited samples. It is important to keep in mind that lymphoblastic lymphomas may be FLI-1 positive.

F- Malignant mixed mullerian tumor (MMMT) of the uterus with neuroectodermal differentiation is extremely rare. In these cases one identifies an intimate blending of the neural tissue with the malignant epithelial and mesenchymal elements of the tumor. It is thought that the neural differentiation in these cases represents neometaplasia of the mullerian elements. The only presence of GFAP immunoreactivity in the neoplastic cells of a MMMT does not imply the presence of neural differentiation.

G- Others

Teratomas with neuroectodermal differentiation are easy to distinguish from PNET as the tumor shows other elements derived from the other germ layers. Pure uterine gliomas are even rarer but the tumors are uniformly composed of cells that resemble glial tissue. The presence of glial tissue in the uterine corpus is a rare but well documented phenomenon. The majority of the patients have a history of therapeutic or spontaneous abortion. The histogenesis of this glial tissue remains speculative. Some authors have tried to explain this finding as a result of implantation of fetal central system tissue in the uterus during the interruption of pregnancy but there is no full agreement among investigators.

Pathogenesis
The origin of uterine PNETs is unclear and several theories have been proposed including: a) mullerian origin, b) development form abnormal migrating neural crest cells, or c) differentiation of native neural elements. The most accepted theory is that PNET evolves from carcinosarcoma that have one-sided neuroectodermal differentiation. Finally, the implantation theory from previous abortions is unlikely as most of these patients are postmenopausal at the time of diagnosis.



Treatment and prognosis
In general, PNETs are associated with a poor prognosis and despite aggressive therapy, the 5-year survival rates are approximately 50%. Patients with uterine PNET are treated with surgery (total abdominal hysterectomy) and lymph node dissection followed by radiation and chemotherapy. Prognosis seems to correlate with stage. Tumors that are confined to the uterus (stage I) have a better prognosis with no evidence of disease after a follow-up interval ranging from months up to 10 years. It appears that survival in younger patients is higher (75%) at > 2 years when compared to older patients.

TAKE HOME MESSAGE:
  • In a reproductive age woman undergoing an endometrial biopsy or a resection of a "fibroid" where the tumor contains a homogeneous population of small blue cells, probably the first two categories of tumors that come to mind include smooth muscle and endometrial stromal tumors. Specific morphologic features associated with these neoplasms should be searched for.

  • Distinction between these two entities may be difficult in some cases and utilization of immunohistochemistry (typically a panel including CD10, desmin and h-caldesmon) helps to definitively categorize the tumor.

  • If the morphology is not exactly right for either a smooth muscle tumor or an endometrial stromal tumor and the immunohistochemical results come back negative (with a good internal positive control), it is important to go one step further going through a list of more unusual tumors that may involve the uterus and that are also composed of "small blue cells" looking for potential characteristic morphologic features and performing immunohistochemical stains or molecular tests that may help to disclose a specific diagnosis.

  • The list of other small blue cell tumors involving the uterus mainly includes PNET, rhabdomyosarcoma and lymphoma. Small cell carcinoma infrequently presents as a "fibroid" but if no clinical history is available this tumor is also in the differential diagnosis.

References
  1. Akbayir O, Gungorduk K, Rafioglu G, et al. Primary primitive neuroectodermal tumor of the uterus: a case report. Arch Gynecol Obstet. 2007.

  2. Baker P, Oliva E. Endometrial stromal tumours of the uterus: a practical approach using conventional morphology and ancillary techniques. J Clin Pathol. 2007;60:235-243.

  3. Baker RJ, Hildebrandt RH, Rouse RV, Hendrickson MR, Longacre TA. Inhibin and CD99 (MIC2) expression in uterine stromal neoplasms with sex-cord-like elements. Hum Pathol. 1999;30:671-9

  4. Bell DA, Woodruff JM, Scully RE. Ependymoma of the broad ligament. A report of two cases. Am J Surg Pathol. 1984;8:203-209.

  5. Blattner JM, Gable P, Quigley MM, et al. Primitive neuroectodermal tumor of the uterus. Gynecol Oncol. 2007;106:419-422.

  6. Borka K, Patai K, Rendek A, et al. Pleomorphic rhabdomyosarcoma of the uterus in a postmenopausal patient. Pathol Oncol Res. 2006;12:102-104.

  7. Campo E, Brunier MN, Merino MJ. Small cell carcinoma of the endometrium with associated ocular paraneoplastic syndrome. Cancer. 1992;69:2283-2288.

  8. Chow SN, Lin MC, Shen J, et al. Analysis of chromosome abnormalities by comparative genomic hybridization in malignant peripheral primitive neuroectodermal tumor of the ovary. Gynecol Oncol. 2004;92:752-760.

  9. Daya D, Lukka H, Clement PB. Primitive neuroectodermal tumors of the uterus: a report of four cases. Hum Pathol. 1992;23:1120-1129.

  10. Ferguson SE, Gerald W, Barakat RR, et al. Clinicopathologic features of rhabdomyosarcoma of gynecologic origin in adults. Am J Surg Pathol. 2007;31:382-389.

  11. Fischer G, Odunsi K, Lele S, et al. Ovarian primary primitive neurectodermal tumor coexisting with endometrioid adenocarcinoma: a case report. Int J Gynecol Pathol. 2006;25:151-154.

  12. Folpe AL, Goldblum JR, Rubin BP, Shehata BM, Liu W, Dei Tos AP, Weiss SW. Morphologic and immunophenotypic diversity in Ewing family tumors: a study of 66 genetically confirmed cases. Am J Surg Pathol 2005;29:1025-33.

  13. Fraggetta F, Magro G, Vasquez E. Primitive neuroectodermal tumour of the uterus with focal cartilaginous differentiation. Histopathology. 1997;30:483-485.

  14. Fukunaga M, Nomura K, Endo Y, et al. Carcinosarcoma of the uterus with extensive neuroectodermal differentiation. Histopathology. 1996;29:565-570.

  15. Gaona-Luviano P, Unda-Franco E, Gonzalez-Jara L, et al. Primitive neuroectodermal tumor of the vagina. Gynecol Oncol. 2003;91:456-458.

  16. Gersell DJ, Duncan DA, Fulling KH. Malignant mixed mullerian tumor of the uterus with neuroectodermal differentiation. Int J Gynecol Pathol. 1989;8:169-178.

  17. Gostout BS, Lindor NM, DiMarco CS, et al. Pelvic primitive neuroectodermal tumor associated with a cluster of small round cell tumors: case report and review of current literature. Gynecol Oncol. 2003;91:247-253.

  18. Gronroos M, Meurman L, Kahra K. Proliferating glia and other heterotopic tissues in the uterus: fetal homografts? Obstet & Gynecol. 1983;61:261-266.

  19. Hendrickson MR, Scheithauer BW. Primitive neuroectodermal tumor of the endometrium:report of two cases, one with electronmicroscopic observations. Int J Gynecol Pathol 1986:5:249-59

  20. Kalifat R, Baup H, Fraitag S, et al. Gliomas of the uterus. Apropos of 3 cases and review of the literature. J Gynecol Obstet Biol Reprod. 1986;15:627-631.

  21. Kleinman GM, Young RH, Scully RE. Primary neuroectodermal tumors of the ovary. A report of 25 cases. Am J Surg Pathol. 1993;17:764-778.

  22. Komuro Y, Mikami M, Sakaiya N, et al. Tumor imprint cytology of ovarian ependymoma. A case report. Cancer. 2001;92:3165-3169.

  23. Lawlor ER, Mathers JA, Bainbridge T, et al. Peripheral primitive neuroectodermal tumors in adults: documentation by molecular analysis. J Clin Oncol. 1998;16:1150-1157.

  24. Lawlor ER, Murphy JI, Sorensen PH, et al. Metastatic primitive neuroectodermal tumour of the ovary: successful treatment with mega-dose chemotherapy followed by peripheral blood progenitor cell rescue. Med Pediat Oncol. 1997;29:308-312.

  25. Liao X, Xin X, Lu X. Primary Ewing's sarcoma-primitive neuroectodermal tumor of the vagina. Gynecol Oncol. 2004;92:684-688.

  26. Malpica A, Moran CA. Primitive neuroectodermal tumor of the cervix: a clinicopathologic and immunohistochemical study of two cases. Annals Diagn Pathol. 2002;6:281-287.

  27. McCluggage WG. Ovarian neoplasms composed of small round cells: a review. Adv Anat Pathol. 2004;11:288-296.

  28. McCluggage WG. Immunohistochemical and functional biomarkers of value in female genital tract lesions. Int J Gynecol Pathol. 2006;25:101-120.

  29. McCluggage WG, Lioe TF, McClelland HR, et al. Rhabdomyosarcoma of the uterus: report of two cases, including one of the spindle cell variant. Int J Gynecol Cancer. 2002;12:128-132.

  30. McCluggage WG, Sumathi VP, Nucci MR, et al. Ewing family of tumours involving the vulva and vagina: report of a series of four cases. J Clin Pathol. 2007;60:674-680.

  31. Mittal S, Sumana G, Gupta M, et al. Primitive neuroectodermal tumor of the uterus: a case report. Int J Gynecol Cancer. 2007;17:524-527.

  32. Musana KA, Raja S, Cangelosi CJ, et al. FDG PET scan in a primitive neuroectodermal tumor. Ann Nucl Med. 2006;20:221-225.

  33. Ng SB, Sirrampalam K, Chuah KL. Primitive neuroectodermal tumours of the uterus: a case report with cytological correlation and review of the literature. Pathol. 2002;34:455-461.

  34. Odunsi K, Olatinwo M, Collins Y, et al. Primary primitive neuroectodermal tumor of the uterus: a report of two cases and review of the literature. Gynecol Oncol. 2004;92:689-696.

  35. Oliva E, Young RH, Amin MB, Clement PB. An immunohistochemical analysis of endometrial stromal and smooth muscle tumors of the uterus: a study of 54 cases emphasizing the importance of using a panel because of overlap in immunoreactivity for individual antibodies. Am J Surg Pathol. 2002; 26:403-12.

  36. Oliva E, Young RH, Clement PB, Bhan AK, Scully RE. Cellular benign mesenchymal tumors of the uterus. A comparative morphologic and immunohistochemical analysis of 33 highly cellular leiomyomas and six endometrial stromal nodules, two frequently confused tumors. Am J Surg Pathol. 1995; 19:757-68.

  37. Ordi J, Stamatakos MD, Tavassoli FA. Pure pleomorphic rhabdomyosarcomas of the uterus. Int J Gynecol Pathol. 1997;16:369-377.

  38. Park JY, Lee S, Kang HJ, et al. Primary Ewing's sarcoma-primitive neuroectodermal tumor of the uterus: a case report and literature review. Gynecol Oncol. 2007;106:427-432.

  39. Pauwels P, Ambros P, Hattinger C, et al. Peripheral primitive neuroectodermal tumour of the cervix. Virchows Arch. 2000;436:68-73.

  40. Peres E, Mattoo TK, Poulik J, et al. Primitive neuroectodermal tumor (PNET) of the uterus in a renal allograft patient: a case report. Pediatr Blood Cancer. 2005;44:283-285.

  41. Rose PG, O'Toole RV, Keyhani-Rofagha S, et al. Malignant peripheral primitive neuroectodermal tumor of the uterus. J Surg Oncol. 1987;35:165-169.

  42. Sato S, Yajima A, Kimura N, et al. Peripheral neuroepithelioma (peripheral primitive neuroectodermal tumor) of the uterine cervix. Tohoku J Exp Med. 1996;180:187-195.

  43. Snijders-Keilholz A, Ewing P, Seynaeve C, et al. Primitive neuroectodermal tumor of the cervix uteri: a case report -- changing concepts in therapy. Gynecol Oncol. 2005;98:516-519.

  44. Sorensen JB, Schultze HR, Madsen EL, et al. Primitive neuroectodermal tumor (PNET) of the uterine cavity. Eur J Obstet Gynecol Reprod Biol. 1998;76:181-184.

  45. Takano T, Akahira J, Moriya T, et al. Primary ependymoma of the ovary: a case report and literature review. Int J Gynecol Cancer. 2005;15:1138-1141.

  46. Tsao AS, Roth LM, Sandler A, et al. Cervical primitive neuroectodermal tumor. Gynecol Oncol. 2001;83:138-142.

  47. Toledo G, Oliva E. Smooth muscle tumors of the uterus: a practical approach. Archiv Pathol & Lab Med (in press)

  48. Ueda G, Yamasaki M. Neuroendocrine carcinoma of the uterus. Curr Top Pathol. 1992;85:309-335.

  49. Vang R, Taubenberger JK, Mannion CM, et al. Primary vulvar and vaginal extraosseous Ewing's sarcoma/peripheral neuroectodermal tumor: diagnostic confirmation with CD99 immunostaining and reverse transcriptase-polymerase chain reaction. Int J Gynecol Pathol. 2000;19:103-109.

  50. Varghese L, Arnesen M, Boente M. Primitive neuroectodermal tumor of the uterus: a case report and review of literature. Int J Gynecol Pathol. 2006;25:373-377.

  51. Ward BS, Hitchcock CL, Keyhani S. Primitive neuroectodermal tumor of the uterus. A case report. Acta Cytologica. 2000;44:667-672.

  52. Whittemore DE, Grondahl RE, Wong K. Primary extraneural myxopapillary ependymoma of the broad ligament. Archiv Pathol & Lab Med. 2005;129:1338-1342.

  53. Wick MR. Immunohistology of neuroendocrine and neuroectodermal tumors. Semin Diagn Pathol. 2000;17:194-203.