Comparative immunohistochemical studies of endometriosis lesions and endometriotic cysts

Size: px

Start display at page:

Download "Comparative immunohistochemical studies of endometriosis lesions and endometriotic cysts"

Alexandrina Chase
6 years ago
Views:

1 FERTILITY AND STERILITY VOL. 78, NO. 4, OCTOBER 2002 Copyright 2002 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Comparative immunohistochemical studies of endometriosis lesions and endometriotic cysts Farr R. Nezhat, M.D., F.A.C.O.G., a and Tamara Kalir, M.D., Ph.D. b The Mount Sinai School of Medicine, New York, New York Received April 8, 2002; revised and accepted June 14, Supported by a grant from the Ovarian Cancer Research Fund. Presented at the 8th World Congress on Endometriosis, San Diego, California, February 24 27, Reprint requests: Farr Nezhat, M.D., Box 1173, Dept. of OB/GYN, The Mount Sinai School of Medicine, 1 Gustav L. Levy Place, New York, New York (FAX: ; farr.nezhat@mssm. edu). a Division of Gynecologic Oncology, Department of Obstetrics, Gynecology, and Reproductive Science. b Division of Gynecologic Pathology, Department of Pathology /02/$22.00 PII S (02) Objective: To compare immunohistochemical staining patterns in noncystic and cystic endometriosis lesions. Design: Experimental. Setting: Archived pathology material in an academic research environment. Patient(s): Endometriosis tissues from the pathology archives including slide tissue sections and blocks. Intervention(s): None; this was a retrospective study. Main Outcome Measure(s): Immunohistochemical staining of the tissues was performed using anti-bcl-2, anti-p53, anti-matrix metalloproteinase IX, and anti-collagen VI antibodies. Staining was qualitatively assessed in terms of extent and intensity. Result(s): p53 showed no staining in both groups. Anti bcl-2 stained 100% (30/30) of endometriosis lesions compared with only 23% (7/30) of endometriotic cysts (P.0001), and anti matrix metalloproteinase IX stained 85% (23/27) of endometriosis lesions and only 39% (14/36) of endometriotic cysts (P.0003). Anti collagen VI, however, stained only 6% (2/35) of endometriosis lesions and 75% (21/28) of endometriotic cysts (P.0001). Conclusion(s): Compared with endometriosis lesions, endometriotic cysts display different expression of proteins with relative overexpression of collagen VI and underexpression of bcl-2 and metalloproteinase IX. This report is the first comparative immunohistochemical study showing these differences. (Fertil Steril 2002;78: by American Society for Reproductive Medicine.) Key Words: Endometriosis, endometriotic cysts, immunohistochemistry, bcl-2, p53, MMP IX, collagen VI Endometriosis is a pathologic condition that affects between 1% and 7% of reproductiveage females and involves growth of endometrial-type tissue outside its normal location in the uterine corpus (1). Currently accepted theories regarding the histogenesis of endometriosis include the implantation of retrograde menstrual tissue, the metaplastic change of coelomic cells, and intravascular spread via lymphatics or intravenous channels (2). Occasionally, endometriosis lesions may be cystic, referred to as endometriomas. Such lesions are most common in the ovary (3 5) and may be a result of implantation of endometrial-type tissue in follicular or luteal cysts (3, 6). We postulated that implantation and maintenance of endometriosis tissue in an already existing cavity (follicle or corpus luteum) would require expression of different genes and proteins than would implantation and maintenance in a noncystic location, prompting our study of comparative immunohistochemical staining of these entities. Endometriomas, in contrast to noncystic endometriosis implants, tend to show foci devoid of epithelial lining (3), and a prominent fibrous wall. These two features prompted our selection of antibodies. Because of the possibility that the absence of epithelial lining is a result of cell death, we decided to study bcl-2 and p53, proteins involved in apoptotic cell death (7). To explore the nature of the fibrous wall, we selected MMP XI and collagen VI proteins, which are secreted into the extracellular (fibrous) matrix. Most studies to date have been performed on endometriosis lesions vs. native uterine endometrium. Jones et al. (8) found elevated bcl-2 expression in stromal cells of endometriosis lesions, compared with paired endometrium samples. Watanabe et al. (9) re- 820

2 ported that the cyclic changes seen in bcl-2 expression in endometrium during the menstrual cycle are absent in endometriosis lesions. To our knowledge, our report is the first comparing solid and cystic endometriosis lesions. MATERIALS AND METHODS Up to 66 cases were studied, approximately evenly composed of ovarian endometriotic cysts and noncystic endometriosis lesions; the latter included lesions from peritoneal and pelvic sites, including noncystic ovarian lesions. The study was approved by the institutional review board. Four-micrometer-thick, unstained sections were cut from archival, routinely processed, formalin-fixed, paraffin-embedded tissues. The cases were retrospectively chosen, based on diagnoses rendered in a clinical setting by gynecologic surgical pathology specialists, including one author (T.K.), who also reviewed the histopathology of each case. The slides were dewaxed in xylene, rehydrated through a graded ethanol series to distilled water, and air dried. Antigen retrieval was performed by incubating the slides in 10-mM citrate buffer, ph 3.0, for 15 minutes in a C water bath, then allowing the hot solution to cool to room temperature (20 minutes). Slides were rinsed in buffer (Chemicon, Temecula, CA). Blocking of endogenous peroxidase activity and protein was performed using blocking reagent (Chemicon secondary detection kit) for 10 minutes, followed by rinsing. Appropriate antibody dilutions were determined using sections of lymph nodes as controls for bcl-2, breast carcinoma for p53, gastric carcinoma for MMP IX, and skin for collagen VI. Optimal final dilutions were 1:50 for bcl-2 antibody (Chemicon), 1:80 for p53 antibody (Calbiochem), 1:8,000 for anti-mmp IX antibody (Chemicon), and 1:9,000 for anti-collagen VI antibody (Chemicon). Antibody solution, in phosphate-buffered saline, ph 7.3, was applied to cover the slide tissue surface and incubated for 2 hours at 37 C for bcl-2, MMP IX, and collagen VI antibodies, and for 2 hours at room temperature for p53 antibody. Slides were then rinsed (Chemicon secondary detection kit), secondary antibodies were applied (Chemicon kit), and the slides were rinsed again. Color development employed streptavidin HRP and DAB (Chemicon kit). After rinsing, slides were counterstained with hematoxylin, dehydrated through a graded ethanol series to xylene, and then coverslipped. The immunohistochemical slides were reviewed on four separate occasions by different combinations of investigators. Golden-brown immunohistochemical staining, when cytoplasmic and membranous, was considered positive for bcl-2, which is located in the plasma membrane and membranes of intracellular organelles. Lymphocytes served as internal positive controls. Golden-brown staining of the cell s cytoplasm and/or extracellular matrix were considered positive for MMP IX and collagen VI. Golden-brown intranuclear staining was scored as positive for p53. TABLE 1 Comparative immunohistochemical staining for bcl-2, matrix metalloproteinase IX, and collagen VI in endometriosis lesions and endometriotic cysts. Antibody Endometriosis, n (%) Extent and intensity of staining were evaluated. Extent assessed roughly how much of the pertinent area in the tissue was stained and was scored as percentages: 0 no staining; 1 1% 25% staining; 2 26% 50% staining; 3 51% 75% staining; 4 75% staining. Intensity assessed the strength of staining and was scored as 1, weak; 2, moderate; and 3, strong or dark golden brown staining. We defined a sample as positive when there was 10% staining of at least weak to moderate (1, 2 ) intensity. Statistical analysis used Fisher s exact test. P.05 was considered statistically significant. RESULTS Group Endometriotic cyst, n (%) Bcl-2 30/30 (100) a 7/30 (23) a MMP IX 26/30 (87) b 14/36 (39) b Collagen VI 2/35 (6) c 22/30 (73) c Data marked with the same superscripts are significantly different (Fisher s exact test). a P.0001 b P.0001 c P.0001 Immunohistochemical staining results are summarized in Table 1. Bcl-2 All 30 endometriosis lesions stained positively with bcl-2, compared with only 7 of 30 endometriotic cysts (Figs. 1 and 2). Staining between the two groups was significantly different (P.0001). Among positively stained cases, generally the glandular component and rarely the stroma showed staining. The case with the lowest number of stained cells contained many glands with secretory change, and these failed to stain with bcl-2. P53 None of the endometriosis lesions or endometriotic cysts stained positively with anti-p53 antibody. Matrix Metalloproteinase IX Twenty-six of 30 endometriosis lesions stained positively, compared with only 14 of 36 endometriotic cysts (Figs. 3 and 4). The results were significantly different (P.0001). In endometriosis, cytoplasmic staining occurred predominantly within the epithelial component. One case showed a decidual reaction FERTILITY & STERILITY 821

FIGURE 1 Positive bcl-2 cytoplasmic staining in the glandular epithelium of a noncystic endometriosis lesion.

In endometriomas, staining was seen in both epithelial and stromal components but was more pronounced in the epithelium.

One case showed simple and complex epithelial hyperplasias, and both stroma and epithelium were strongly stained.

Staining results were significantly different (P.0001). Positive staining tended to occur in the epithelium, and in some cases in the stromal cells.

3 FIGURE 1 Positive bcl-2 cytoplasmic staining in the glandular epithelium of a noncystic endometriosis lesion. FIGURE 2 Negative staining for bcl-2 in an endometriotic cyst. of the stroma, and these cells were strongly positive. In endometriomas, staining was seen in both epithelial and stromal components but was more pronounced in the epithelium. Metaplastic cells (ciliated, clear cell, tubal) tended to stain. One case showed simple and complex epithelial hyperplasias, and both stroma and epithelium were strongly stained. Collagen VI Only 2 of 35 cases of endometriosis stained positively with collagen VI, compared with 22 of 30 endometriotic cysts (Figs. 5 and 6). Staining results were significantly different (P.0001). Positive staining tended to occur in the epithelium, and in some cases in the stromal cells. Occasionally occurring collagenous areas were positively stained. DISCUSSION Endometriosis lesions stained differently than endometriotic cysts for three of the four proteins studied. FIGURE 3 Positive cytoplasmic staining for MMP-9 in glandular epithelium of a noncystic endometriosis lesion. FIGURE 4 Sparse MMP-9 staining in epithelial cells lining an endometriotic cyst. 822 Nezhat and Kalir Endometriosis vs. endometriomas Vol. 78, No. 4, October 2002

It is believed that the mutated protein is the one more likely to be stained immunohistochemically if present, because of its longer half-life (10).

4 FIGURE 5 Negative staining for collagen VI in a noncystic endometriosis lesion. Note adjacent positively stained fibrous tissue (right). FIGURE 6 Positive collagen VI cytoplasmic staining in epithelial cells lining an endometriotic cyst. P53 The antibody used detects both wild-type and mutated p53. It is believed that the mutated protein is the one more likely to be stained immunohistochemically if present, because of its longer half-life (10). Mutated p53 is held to be a marker of malignancy, and so it is not surprising that we found a lack of staining of the endometriosis lesions and endometriotic cysts. Bcl-2 This protein is generally held to be a negative regulator of apoptotic cell death due to DNA damage. Decreased expression might allow for apoptosis to occur; however, we have not found a notable number of apoptotic bodies in our cases on examination of the hematoxylin and eosin stained slides. Thus decreased bcl-2 protein in endometriomas may be due to other causes. Additional roles for bcl-2 have been proposed in the literature, including involvement in glandular cell proliferation and differentiation and an antimitotic role (9, 11). It has been suggested that bcl-2 expression in endometriosis lesions allows for survival of the tissue in ectopic sites (8). The different bcl-2 staining between noncystic and cystic endometriosis suggests that these two entities may use different pathways for their development and/or maintenance and is in keeping with the idea that endometriomas may result from implantation of endometrial-type tissue in a follicle or corpus luteum. Alternatively, differential expression may not be causative in, but an effect of the different microenvironments of the two entities; for example, [1] location within the ovary, where hormone levels may be higher than in extraovarian sites, or [2] endometriotic cyst lining, in comparison to noncystic endometriosis lesions, may be subject to pressure atrophy by the cyst fluid. Regarding ovarian location, we found that of a total of 30 noncystic endometriosis lesions, 14 were ovarian, and the other 16 were from extraovarian sites, including pelvic and abdominal peritoneum, bladder, skin, uterosacral ligament, omentum, vagina, colon, and appendix. The staining among the two groups (ovarian vs. extraovarian endometriosis) was similar and different from cystic ovarian endometriosis lesions. Thus, location within the ovary does not explain the staining differences between cystic and noncystic endometriosis lesions. Regarding the possibility of pressure atrophy from cyst fluid, we found that many of our negatively stained endometriomas did not exhibit atrophy. They showed a thick internal cyst lining composed of stroma and pseudostratified epithelial cells with columnar (not flattened) cytoplasm, sometimes with metaplastic changes. One negatively stained sample showed cystic (simple) hyperplasia. A possibility that we cannot eliminate is that constituents of the cyst fluid effect gene expression in the lining cells. Matrix Metalloproteinase IX Matrix metalloproteinases are a family of proteins that degrade various components of the extracellular matrix (12). Matrix metalloproteinase IX degrades type IV collagen, a major component of the basement membrane, and has been implicated in tumor cell invasion and metastasis (12) and in the pathogenesis of endometriosis (2). Our findings of relatively strong protein expression in endometriosis lesions lend support to the idea that extracellular matrix degradation may be important for establishing endometriosis lesions, and possibly for their maintenance. Endometriomas showed relative underexpression, compared with noncystic endometri- FERTILITY & STERILITY 823

5 osis lesions, which suggests that endometriotic cysts do not require extracellular matrix degradation for their establishment and/or maintenance. This could occur if the endometriosis were to grow along an already established space. This finding also lends support to the idea (3) that endometriotic cysts are a result of growth of endometriosis tissue along the internal surface of follicular or corpus luteal cysts. Collagen VI Adherence to and invasion of the extracellular matrix have been implicated in the pathogenesis of endometriosis (13 16). Among matrix components, collagens I, III, and IV reportedly have similar distributions in ectopic endometriosis implants and intrauterine endometrium (17). As far as we are aware, collagen VI has not been studied in endometriosis and endometrium. Collagen VI is present in a number of adult connective tissues (18) and has been found to be overly expressed in ovarian cancer cells (19). The relative overexpression of this protein in endometriomas in comparison to noncystic endometriosis lesions is interesting. This difference may be attributable to the fact that endometriotic cysts contain fibrous tissue in their wall, of which collagen VI may be a major component. Alternatively, enzymes in control of collagen VI degradation may be inhibited in this microenvironment. Inflammation, more commonly seen histologically in endometriomas, may also contribute to the accumulation of this protein. In summary, ovarian endometriotic cysts stained differently from noncystic endometriosis implants, suggesting that different genes are expressed in the development and maintenance of these two entities. Alternatively, protein expression may be a result of the different microenvironments, in other words, endometriomas have adjacent cyst fluid, whereas noncystic endometriosis lesions do not. Future studies with additional markers may lend additional insight. Acknowledgments: The authors thank John Mandeli, Ph.D., Department of Biomathematics, The Mount Sinai School of Medicine, for statistical analysis. References 1. Jones RK, Bulmer JN, Searle RF. Immunohistochemical characterization of stromal leukocytes in ovarian endometriosis: comparison of eutopic and ectopic endometrium with normal endometrium. Fertil Steril 1996;66: Witz CA. Current concepts in the pathogenesis of endometriosis. Clin Obstet Gynecol 1999;42: Nezhat F, Nezhat C, Allan CJ, Metzger DA, Sears DL. Clinical and histologic classification of endometriomas. Implications for a mechanism of pathogenesis. J Reprod Med 1992;37: Wolf Y, Haddad R, Werbin N, Skornick Y, Kaplan O. Endometriosis in abdominal scars: a diagnostic pitfall. Am Surg 1996;62: Healy JT, Wilkinson NW, Sawyer M. Abdominal wall endometrioma in a laparoscopic trocar tract: a case report. Am Surg 1995;61: Nezhat C, Nezhat F, Nezhat C, Admon D. Treatment of ovarian endometriosis. In: Nezhat CR, Berger GS, Nezhat FR, Buttram VC Jr, Nezhat CH, eds. Endometriosis. Advanced management and surgical techniques. New York: Springer-Verlag, 1995: Marone M, Scambia G, Mozzetti S, Ferrandina G, Iacovella S, De Pasqua A, et al. Bcl-2, bcl-xl, and bcl-xs expression in normal and neoplastic ovarian tissues. Clin Cancer Res 1998;4: Jones RK, Searle RF, Bulmer JN. Apoptosis and bcl-2 expression in normal human endometrium, endometriosis and adenomyosis. Hum Reprod 1998;13: Watanabe H, Kanzaki H, Narukawa S, Inoue T, Katsuragawa H, Kaneko Y, et al. Bcl-2 and Fas expression in eutopic and ectopic human endometrium during the menstrual cycle in relation to endometrial cell apoptosis. Am J Obstet Gynecol 1997;176: Geisler JP, Geisler HE, Miller GA, Wiemann MC, Zhou Z, Crabtree W. p53 and bcl-2 in epithelial ovarian carcinoma: their value as prognostic indicators at a median follow-up of 60 months. Gynecol Oncol 2000; 77: Crescenzi E, Palumbo G. Bcl-2 exerts a prb-mediated cell cycle inhibitory function in HEC1B endometrial carcinoma cells. Gynecol Oncol 2001;81: Huang L-W, Garrett AP, Bell DA, Welch WR, Berkowitz RS, Mok SC. Differential expression of matrix metalloproteinase-9 and tissue inhibitor of metalloproteinase-1 protein and mrna in epithelial ovarian tumors. Gynecol Oncol 2000;77: Koks CA, Groothuis PG, Dunselman GA, degroeij AF, Evers JL. Adhesion of menstrual endometrium to extracellular matrix: the possible role of integrin alpha (6)beta(1) and laminin interaction. Mol Hum Reprod 2000;6: Starzinski-Powitz A, Gaetje R, Zeitvogel A, Kotzian S, Handrow- Metzmacher H, Herrmann G, et al. Tracing cellular and molecular mechanisms involved in endometriosis. Hum Reprod Update 1998;4: Gaetje R, Kotzian S, Herrmann G, Baumann R, Starzinski-Powitz A. Invasiveness of endometriotic cells in vitro. Lancet 1995;346: Spuijbroek MD, Dunselman GA, Menheere PP, Evers JL. Early endometriosis invades the extracellular matrix. Fertil Steril 1992;58: Stovall DW, Anners JA, Halme J. Immunohistochemical detection of type I, III, and IV collagen in endometriosis implants. Fertil Steril 1992;57: Hessle H, Engvall E. Type VI collagen. Studies on its localization, structure, and biosynthetic form with monoclonal antibodies. J Biol Chem 1984;259: Ismail RS, Baldwin RL, Fang J, Browning D, Kalian BY, Gasson JC, et al. Differential gene expression between normal and tumor-derived ovarian epithelial cells. Cancer Res 2000;60: Nezhat and Kalir Endometriosis vs. endometriomas Vol. 78, No. 4, October 2002

Cancer arising from Endometriosis and Its Clinical implications

Cancer arising from Endometriosis and Its Clinical implications 1) Nezhat F, Cohen C, Rahaman J, Gretz H, Cole P, Kalir T. Comparative immunohistochemical studies of bcl-2 and p53 proteins in benign