February 13, 2008
Immunohistochemistry in Gynecologic Pathology
By Deavers, Michael T
Although most diagnoses in gynecologic pathology are established through the examination of routine hematoxylin eosin-stained slides, immunohistochemical studies can be of assistance in some cases. The following case studies are used to illustrate the value of some relatively new markers in the diagnosis of gynecologic neoplasms and to highlight potential diagnostic pitfalls. (Arch Pathol Lab Med. 2008;132:175-180)CASE 1
The patient is a 59-year-old woman, nonsmoker, who presented 11 months prior with postmenopausal spotting. Following an en- dometrial biopsy, she underwent a total abdominal hysterectomy and bilateral salpingo-oophorectomy (TAH-BSO) and lymph node dissection for endometrioid endometrial adenocarcinoma, grade 2. Histologic examination showed that the tumor invaded 9 mm into an 18-mm-thick myometrium and involved the cervical stroma (Figure 1). There was lymphvascular involvement, but the pelvic and para-aortic lymph nodes were negative for metastatic carcinoma. The patient received adjuvant brachytherapy.
Ten months after the TAH-BSO, her serum CA 125 was 41 U/ mL. A computed tomography scan of the chest demonstrated multiple bilateral pulmonary nodules, and a computed tomog-raphy scan of the abdomen and pelvis revealed changes consistent with carcinomatosis. A diagnosis of adenocarcinoma was made on material obtained from a computed tomography-guided fine-needle aspiration of the lung (Figure 2). The tumor cells were found to be positive for cytokeratin 7 and thyroid transcription factor 1 (TTF-1) (Figure 3).
Initially, because of the TTF-1 expression by tumor cells, a new pulmonary primary was favored. However, the relatively short interval since the patient's TAH-BSO, the bilateral and multiple nature of the pulmonary nodules, and the intra-abdominal disease were more consistent with recurrent endometrial carcinoma. Subsequently, sections of the uterine tumor were submitted for immunohistochemistry. The endometrial carcinoma was found to be diffusely positive for TTF-1 and was also positive for estrogen receptor (ER) and progesterone receptor (PR) (Figure 4). Chrom- ogranin and synaptophysin stains were negative. The diagnosis of the pulmonary aspirate was revised to metastatic endometrial carcinoma.
Thyroid transcription factor 1 is a nuclear protein of the NKx2 family of homeodomain transcription factors. Thyroid transcription factor 1 was originally found in follicular epithelial cells of the thyroid and later in type II pneumocytes and Clara cells of the lung. In the thyroid, it regulates the expression of thyroid peroxidase, thyro-globulin, and the thyroid-stimulating hormone receptor. In the lung, it regulates the expression of surfactant protein A, surfactant protein B, surfactant protein C, and Clara cell secretory protein. Other normal tissues expressing TTF-1 include parafollicular cells, parathyroid gland, anterior pituitary, and fetal basal forebrain.
Nuclear expression of TTF-1 has been found to be relatively specific and sensitive for tumors originating in the thyroid and lung. Follicular adenomas, follicular carcinomas, papillary carcinomas (including follicular variants), medullary carcinomas, and some anaplastic thyroid carcinomas are positive for TTF-1. Most pulmonary adeno-carcinomas (decreased expression in mucinous vs non- mucinous), small cell carcinomas, and carcinoid tumors are also positive.1,2 Additionally, TTF-1 is expressed by some large cell carcinomas of the lung, both neuroendo-crine and nonneuroendocrine.
The specificity of TTF-1 for thyroid and lung neoplasms, however, is not 100%. Small cell carcinomas from various sites, including the cervix, bladder, gastrointestinal tract, and prostate, are positive in 15% to 33% of cases.3 Other tumors with reports of occasional TTF- 1 expression include ependymoma (2/27),4 breast carcinoma (A. Sahin, MD, oral communication, August 2006), gastric carcinoma (1/66),5 and colonic adenocarcinoma (4/41).6 Of interest, the study of the colonic adenocarcinomas found that the SPT24 clone of anti-TTF-1 is more sensitive than the 8G7G3/1 clone. For the diagnosis of hepatocellular carcinoma, cytoplasmic reactivity for TTF-1 (rather than nuclear expression) has been reported to be useful.7
In the gynecologic tract, endocervical and endometrial carcinomas occasionally express nuclear TTF-1. In the ovary, struma ovarii, of course, expresses TTF-1. However, 2 studies have also reported positivity of ovarian adenocar-cinomas. One, a study of effusion and fine-needle aspiration specimens, found focal weak reactivity in 1 of 16 cases of metastatic ovarian carcinoma; the histotype was not specified.8 A larger study using tissue microarrays examined 166 ovarian tumors of various histologic types; 2 of the tumors were positive. One, a stage III high-grade mixed serous and endometrioid adenocarcinoma, had strong staining in all 4 microarray cores; the other was a stage III high-grade serous carcinoma with strong staining in 1 of 3 cores. Neither of the patients had evidence of a pulmonary lesion.9
In the cervix and endometrium, immunoreactivity for TTF-1 has been reported in 1 of 3 cervical small cell carcinomas3 and 1 of 8 endometrial adenocarcinomas.5 At M. D. Anderson, we recently studied TTF-1 expression in 28 endocervical adenocarcinomas (9 well, 12 moderately, and 6 poorly differentiated and 1 adenosquamous) and 31 en-dometrial endometrioid adenocarcinomas (11 grade I, 8 grade II, and 12 grade III); none of the tumors had a neu-roendocrine component. One (4%) of the endocervical ad-enocarcinomas was positive, with a diffuse distribution of staining (Figures 5 and 6). Thyroid transcription factor 1 immunoreactivity was more frequent in the endometrial tumors, being present in 5 cases (16%), and ranged from focal to diffuse in distribution. There was no correlation between the degree of differentiation and TTF-1 expression, and no distinguishing histologic features of the positive tumors were noted.
Thyroid transcription factor 1 is a sensitive and relatively specific marker of pulmonary and thyroidal tumors. However, a positive result does not automatically equate to origin from these sites, as a small percentage of carcinomas from other sites such as the cervix, endometrium, and ovary are also positive. Thyroid transcription factor 1, as with any other marker, should be interpreted in the context of the clinical setting, radiologic findings, histo-logic features, and other markers.
The patient is a 70-year-old woman who presented to her primary care physician with shortness of breath and a cough. The patient's medical history was significant for a TAH-BSO 15 years prior for uterine ''fibroids.'' A chest x-ray and a subsequent computed tomography scan of the chest demonstrated multiple pulmonary and pleural nodules. Later she was noted to have left hip and leg pain. A magnetic resonance imaging scan revealed a lytic lesion that had destroyed part of the femur. She underwent resection of the distal femur and reconstruction with an endo-prosthesis.
The tumor involved the distal diaphysis and metaphysis and extended through the cortex into adjacent soft tissue. The intra- medullary component measured 21 x 4 x 4 cm, and the soft tissue component was 12 x 7 x 5 cm. Histologically, it was composed of intersecting fascicles of spindle cells with eosinophilic cytoplasm (Figure 7). The nuclei were elongated and oval with mild to moderate atypia; there were 3 mitoses per 10 high-power fields. Focal coagulative tumor cell necrosis was present. The tumor infiltrated medullary bony trabeculae but did not produce matrix itself. It was diffusely positive for smooth muscle actin, desmin, and Wilms tumor 1 (WT-1) (Figure 8). Estrogen receptor was expressed by 80% of the tumor cells (Figure 9); PR was negative. A diagnosis of leiomyosarcoma was made.
Leiomyosarcoma involving bone, whether primary or secondary, is unusual. Approximately 120 cases of primary leiomyosarcoma of the bone have been reported.10 The patients are generally middle-aged or older, men and women are equally affected, and the long bones of the extremities are most frequently involved. However, before a leiomyosarcoma can be accepted as arising from the bone, metastatic disease must be ruled out. In these cases, the uterus is one of the most common primary sites, with metastases occurring up to 19 years after hysterectomy, and a careful rereview of all previous myomectomy and hysterectomy specimens is recommended.
Unfortunately, the slides and blocks from the hysterectomy specimen in this case had been discarded, which is illustrative of the problems sometimes encountered with extrauterine smooth muscle tumors. Often the patient's gynecologic history is vague or unknown. Additionally, although the criteria for malignancy differ, the histologic features of uterine and extrauterine smooth muscle tumors are the same. Immunohistochemistry may offer some guidance in these cases.11-19
Most uterine smooth muscle tumors express hormone receptors. Virtually all leiomyomas are positive for ER and PR, whereas 36% to 87% of leiomyosarcomas are positive for ER and 43% to 80% are positive for PR. This is in contrast to extrauterine smooth muscle tumors. Only a few extrauterine leiomyomas are reported to be ER and PR positive, and only 0% to 25% and 0% to 13% of extrauter-ine leiomyosarcomas are reported to express ER and PR, respectively. Additionally, although uterine leiomyosar-comas often have a diffuse distribution of receptors, ER and PR are usually only focally expressed in extrauterine primary tumors.13,15,16,17,19 In addition to their potential utility in helping to distinguish uterine from extrauterine leiomyosarcomas, ER and PR have also been investigated in regard to the diagnosis and prognosis of uterine smooth muscle tumors. Three studies have suggested that leiomyosarcomas, as a group, have decreased expression of hormone receptors in comparison with benign smooth muscle tumors and smooth muscle tumors of uncertain malignant potential. However, given the relatively limited initial experience and the overlap in staining between these categories of smooth muscle tumors, great caution should be exercised in the use of ER and PR expression for the diagnosis of malignant potential in individual cases. No significant correlation has been found between ER and PR expression and recurrence or survival.11,12,18
WT-1, the Wilms tumor gene involved in the development of the genitourinary system, has been noted recently to be present in uterine cellular leiomyomas.11 At M. D. Anderson, we undertook a study examining WT-1 expression in a wide range of mullerian and nonmullerian smooth muscle tumors. All of the gynecologic leiomyomas and smooth muscle tumors of uncertain malignant potential and approximately 50% of the leiomyosarcomas (including metastases) were positive for WT-1 (nuclear ex-pression). However, none of the nongynecologic smooth muscle tumors (including leiomyosarcomas) were positive.14
When faced with an extrauterine smooth muscle tumor of uncertain primary, the patient's history and a review of prior specimens can be helpful for resolving questions of the tumor's origin. Additionally, immunohistochemical staining for ER, PR, and WT-1 may be of assistance. In this case, because of the ER and WT-1 expression, the leio-myosarcoma was considered most consistent with metastasis from a prior uterine primary.
The patient is a 45-year-old woman who presented to her gynecologist with a complaint of abnormal uterine bleeding. She had been treated in the past for cervical dysplasia. The uterus was not enlarged and no abnormality of the cervix was visualized on examination. An endometrial biopsy was obtained.
On histologic examination, the specimen contained atypical papillary and glandular fragments consistent with adenocarci-noma (Figure 10). No normal tissue was seen. The carcinoma was diffusely positive for p16 (Figure 11), there was only focal staining for vimentin, and staining for ER and carcinoembryonic antigen was negative. Although there was some overlap in the characteristics of this profile, the diffuse staining for p16, negative ER staining, and minimal vimentin supported an endocervical primary. The patient underwent a TAH-BSO and was found to have endometrial serous carcinoma.
p16 (cyclin-dependent kinase inhibitor 4) is a tumor suppressor protein that binds to cyclin-cdk4/6 complexes. It blocks their kinase activity and inhibits progression to the S phase of the cell cycle.20-24 In the gynecologic tract, p16 overexpression has been found to be a surrogate marker of high-risk human papillomavirus- associated neoplasia.20 This is characterized by a pattern of diffuse strong nuclear and cytoplasmic positivity.
In the vulva, vulvar intraepithelial neoplasia 3 of the usual and basaloid types, and the squamous carcinomas associated with them, diffusely express p16 (40/40).24 However, differentiated vulvar intraepithelial neoplasia, or vulvar intraepithelial neoplasia simplex, which is generally found in older women, and its associated invasive squamous carcinoma are negative for p16 (0/11).
In the cervix, diffuse strong p16 expression correlates well with high-grade squamous dysplasia (80%-100%), although some low-grade squamous dysplasias may stain (Figure 12). Invasive squamous carcinomas of the cervix diffusely express p16 at the same rate as the high-grade dysplasias. Together with Ki-67, p16 also can be used to help confirm the presence of adenocarcinoma in situ, and the usual type of endocervical adenocarcinoma expresses p16 in most cases (>90%)22 (Figure 13). An exception to the expression of p16 by endocervical adenocarcinomas appears to be minimal deviation adenocarcinoma. Many minimal deviation adenocarcinomas do not appear to be associated with high-risk human papillomavirus, and accordingly only a minority are positive for p16 (30%).22
Because the majority of endocervical adenocarcinoma in situ and invasive adenocarcinoma cases diffusely express p16, this marker is now used frequently to distinguish endocervical glandular neoplasia from benign conditions and is incorporated into a panel of markers to help distinguish endocervical adenocarcinomas from endometrial primaries.21 In contrast to the strong diffuse nuclear and cytoplasmic staining of endocervical adenocarcinomas, en-dometrial endometrioid adenocarcinomas exhibit only patchy staining. A potential pitfall, however, is that in small specimens an endometrial endometrioid adenocar-cinoma may seem to overexpress p16 and therefore mimic an endocervical primary (Figure 14). Another potential pitfall is the diffuse expression of p16 by serous and clear cell endometrial carcinomas.
At M. D. Anderson, we recently studied the expression of p16 in 11 cases of uterine serous carcinoma and 10 cases of ovarian high- grade serous carcinoma. All of the uterine cases exhibited diffuse strong p16 staining. In fact, p16 staining was more prominent than p53. There was also diffuse staining for p16 in half of the ovarian serous carcinoma cases. In clear cell carcinomas of the endometrium, approximately one third of the tumors may be diffusely positive, and in ovarian clear cell carcinomas approximately 75% overexpress p16.
In summary, p16 may be useful in the diagnosis of high-risk human papillomavirus-related neoplasia of the gynecologic tract, including high-grade squamous dyspla-sia, invasive squamous carcinoma, endocervical adenocar-cinoma in situ, and invasive endocervical adenocarcinoma in younger women, and to help confirm vulvar or cervical metastases at distant sites. It may also be used as part of a panel to distinguish endocervical from endometrial primaries. However, a potential pitfall in these instances is the diffuse expression of p16 by non-human papillomavirus related high-grade endometrial carcinomas. p16 should be used with caution, if at all, in a panel for cases of poorly differentiated or high-grade carcinoma.
1. Khoor A, Whitsett JA, Stahlman MT, Olson SJ, Cagle PT. Utility of surfactant protein B precursor and thyroid transcription factor 1 in differentiating adenocar-cinoma of the lung from malignant mesothelioma. Hum Pathol. 1999;30:695- 700.
2. Byrd-Gloster AL, Khoor A, Glass LF, et al. Differential expression of thyroid transcription factor 1 in small cell lung carcinoma and Merkel cell tumor. Hum Pathol. 2000;31:58-62.
3. Ordonez NG. Value of thyroid transcription factor-1 immunostaining in distinguishing small cell lung carcinomas from other small cell carcinomas. Am J Surg Pathol. 2000;24:1217-1223.
4. Zamecnik J, Chanova M, Kodet R. Expression of thyroid transcription factor 1 in primary brain tumors. J Clin Pathol. 2004;57:1111-1113.
5. Bejarano PA, Baugham RP, Biddinger PW, et al. Surfactant proteins and thyroid transcription factor-1 in pulmonary and breast carcinomas. Mod Pathol. 1996;9:445-452.
6. Comperat E, Zhang F, Perrotin C, et al. Variable sensitivity and specificity of TTF-1 antibodies in lung metastatic adenocarcinoma of colorectal origin. Mod Pathol. 2005;18:1371-1376.
7. Wieczorek TJ, Pinkus JL, Glickman JN, Pinkus GS. Comparison of thyroid transcription factor-1 and hepatocyte antigen immunohistochemical analysis in the differential diagnosis of hepatocellular carcinoma metastatic adenocarcino-ma, renal cell carcinoma, and adrenal cortical carcinoma. Am J Clin Pathol. 2002;118:911-921.
8. Hecht JL, Pinkus JL, Weinstein LJ, Pinkus GS. The value of thyroid transaction factor-1 in cytologic preparations as a marker for metastatic adenocarcinoma of lung origin. Am J Clin Pathol. 2001;116:483-488.
9. Graham AD,Williams ARW, Salter DM. TTF-1 expression in primary ovarian epithelial neoplasia. Histopathology. 2006;48:746-765.
10. Fornasier VL, Paley D. Leiomyosarcoma in bone: primary or secondary? A case report and review of the literature. Skeletal Radiol. 1983;10:147-153.
11. Agoff SN, Grieco VS, Garcia Gown AM. Immunohistochemical distinction of endometrial stromal sarcoma and cellular leiomyoma. Appl Immunohistochem Mol Morphol. 2001;9:164-169.
12. Bodner K, Bodner-Adler B, Kimberger O, Czerwenka K, Leodolter S, May-erhofer K. Estrogen and progesterone receptor expression in patients with uterine leiomyosarcoma and correlation with different clinicopathological parameters. Anticancer Res. 2003;23:729-732.
13. Bodner K, Bodner-Adler B, Kimberger O, Czerwenka K, Mayerhofer K. Estrogen and progesterone receptor expression in patients with uterine smooth muscle tumors. Fertil Steril. 2004;81:1062-1066.
14. Deavers MT, Silva EG, Euscher ED, Liu J, Broaddus RB, Malpica A. WT1 expression may differentiate mullerian from non-mullerian smooth muscle tumors. Mod Pathol. 2006;19(suppl 1):176A.
15. Kelly TW, Borden EC, Goldblum JR. Estrogen and progesterone receptor expression in uterine and extrauterine leiomyosarcomas: an immunohistochem-ical study. Appl Immunohistochem Mol Morphol. 2004;12:338-341.
16. Paal E, Miettinen M. Retroperitoneal leiomyomas: a clinicopathologic and immunohistochemical study of 56 cases with a comparison to retroperitoneal leiomyosarcomas. Am J Surg Pathol. 2001;25:1355-1363.
17. Rao UN, Finkelstein SD, Jones MW. Comparative immunohistochemical and molecular analysis of uterine and extrauterine leiomyosarcomas. Mod Pathol. 1999;12:1001-1009. 18. Watanabe K, Suzuki T. Uterine leiomyoma versus leiomyosarcoma: a new attempt at differential diagnosis based on their cellular characteristics. Histopa-thology. 2006;48:563-568.
19. Zhai YL, Kobayashi Y, Mori A, et al. Expression of steroid receptors, Ki-67, and p53 in uterine leiomyosarcomas. Int J Gynecol Pathol. 1999;18:20-28.
20. Keating JT, Ince T, Crum CP. Surrogate biomarkers of HPV infection in cervical neoplasia and diagnosis. Adv Anat Pathol. 2001;8:83-92.
21. McCluggage WG, Jenkins D. Immunohistochemical staining with p16 may assist in the distinction between endometrial and endocervical adenocarcinoma. Int J Gynecol Pathol. 2003;22:231-235.
22. Mikami Y, Kiyokawa T, Hata S, et al. Gastrointestinal immunophenotype in adenocarcinomas of the uterine cervix and related glandular lesions: a possible link between lobular endocervical glandular hyperplasia/pyloric gland metaplasia and adenoma malignum. Mod Pathol. 2004;17:962-972.
23. Reid-Nicholson M, Iyengar P, Hummer AJ, et al. Immunophenotypic diversity of endometrial adenocarcinomas: implications for differential diagnosis. Mod Pathol. 2006;19:1091- 1100.
24. Santos M, Montagut C, Mellado B, et al. Immunohistochemical staining for p16 and p53 in premalignant and malignant epithelial lesions of the vulva. Int J Gynecol Pathol. 2004;23:206-214.
Michael T. Deavers, MD
Accepted for publication October 10, 2007.
From the Department of Pathology, The University of Texas M. D. Anderson Cancer Center, Houston.
The author has no relevant financial interest in the products or companies described in this article.
Presented at the 86th Annual Meeting of the Texas Society of Pathologists, Fort Worth, Tex, January 26-28, 2007.
Reprints: Michael T. Deavers, MD, Department of Pathology, Box 85, The University of Texas M. D. Anderson Cancer Center, 1515 Hol- combe Blvd, Houston, TX 77030 (e-mail: [email protected] org).
Copyright College of American Pathologists Feb 2008
(c) 2008 Archives of Pathology & Laboratory Medicine. Provided by ProQuest Information and Learning. All rights Reserved.