- Email: [email protected]
Altered Expression of CD44 and Variant Isoforms in Human Adenocarcinoma of the Endocervix during Progression
Gynecologic Oncology 75, 84 –90 (1999) Article ID gyno.1999.5517, available online at http://www.idealibrary.com on
Altered Expression of CD44 and Variant Isoforms in Human Adenocarcinoma of the Endocervix during Progression 1 Di Lu, M.D., Ph.D.,* Ossama Tawfik, M.D., Ph.D.,* ,† Cooley Pantazis, M.D.,‡ Wendy Hobart, M.D.,† Julia Chapman, M.D.,† and Kenneth Iczkowski, M.D.§ *Department of Pathology and Laboratory Medicine and †Department of Obstetrics and Gynecology, University of Kansas Medical Center, Kansas City, Kansas 66160; ‡Department of Pathology, University of Florida, Gainesville, Florida 32601, and Department of Pathology, Munroe Regional Medical Center, Ocala, Florida 34478; and §Department of Pathology, Reading Hospital and Medical Center, Reading, Pennsylvania 19612 Received November 18, 1998
INTRODUCTION
Objective. Altered expression of the CD44 family of cell adhesion molecules has been associated with tumor progression and metastasis. The aim of this study was to investigate the expression of the gene products of CD44 standard (CD44s) and several alternatively spliced variants (CD44v4, v6, v7, and v9) in adenocarcinoma of the endocervix and to correlate the degree of their expression with disease progression. Methods. Immunohistochemical staining for CD44s and CD44v4, v6, v7, and v9 was performed on formalin-fixed, paraffin-embedded endocervical specimens. Seventeen cases of adenocarcinoma in situ (AIS) and 22 cases of invasive adenocarcinoma of the endocervix were included in this study, and the immunoreactivity was compared with that of normal endocervical epithelium. Results. (1) In the normal endocervical mucosa, immunoreactivity for CD44s and the splice variants was lacking or was confined to only the basal portion of the glandular epithelium along the basement membrane; (2) CD44s was diffusely expressed along the entire cytoplasmic membrane, including the luminal surface of the tumorous glands in 94% of AIS and 95% of invasive adenocarcinomas; (3) a significantly stronger expression of CD44s was observed in invasive adenocarcinomas than in AIS; (4) in contrast to all other splice variants, CD44v9 demonstrated an increased expression in nearly all in situ and invasive lesions compared to the normal tissue; (5) CD44v4 and v6 were expressed only in a small proportion of invasive adenocarcinomas and were near totally absent in the in situ lesions; and (6) CD44 v7 was totally absent in all normal, in situ, and invasive lesions studied. Conclusions. It appears that neoplastic transformation of endocervical epithelium is associated with qualitative and quantitative changes in the expression of CD44 standard molecule and some CD44 splice variants. © 1999 Academic Press Key Words: CD44; variant isoforms 4, 6, 7, and 9; endocervical adenocarcinoma.
Adenocarcinoma of the endocervix is a relatively common gynecologic malignancy that constitutes 7 to 26% of primary cervical carcinomas [1–5]. As is well known for squamous epithelium, it appears that endocervical epithelium progresses through a continuum of morphologic abnormalities beginning with endocervical dysplasia/adenocarcinoma in situ and culminating in the classic invasive adenocarcinoma. Clinically, the majority of preinvasive lesions and up to 20% of invasive adenocarcinomas are asymptomatic and are usually invisible on routine colposcopic examination [6, 7]. Otherwise, patients present with symptoms such as abnormal uterine bleeding, vaginal discharge, or pain. Epidemiologically, endocervical lesions have shown a definite age progression with increasing lesional severity. Cervical glandular lesions, particularly adenocarcinoma in situ, are associated with squamous intraepithelial lesion and human papillomavirus infection. However, the pathogenesis of endocervical adenocarcinoma is not well understood. Tumor development is usually associated with dysfunctional expression or mutation of various genes, including oncogene overexpression, aberrant tumor suppressor gene expression, or deletion and the inappropriate expression of a variety of cytokines and growth factors. CD44 is a family of transmembrane glycoproteins involved in cell– cell and cell– extracellular matrix interaction [8]. The CD44 gene, which maps to chromosome 11, contains at least 21 exons spanning about 60 kb and can be subdivided into five structural domains. The standard CD44 molecule plays an important role in lymphocyte homing, tissue regeneration, and tumor cell dissemination [8 –12]. At least 25 different CD44 variant isoforms can be derived from alternative splicing and more heterogeneity results from extensive posttranslational modification of the protein products. The normal functions of CD44 variants are entirely unknown, however. Altered expression of both standard molecule and certain splice variants of CD44 has been linked with invasive and metastatic behaviors of several human malignancies such as colorectal carcinoma
1 Part of this article was presented in poster form at the Annual Fall Meeting of the American Society of Clinical Pathologists, Philadelphia, September 1997.
0090-8258/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
84
ALTERED EXPRESSION OF CD44 IN ADENOCARCINOMA
[13, 14], prostatic adenocarcinoma [15–17], bladder cancer [18, 19], renal cell carcinoma [20, 21], brain tumors [22, 23], skin cancers [24 –26], and breast cancer [27–29]. In most studies, increased expression of CD44 standard and/or variants correlated with progression and metastatic behavior of malignancies. Decreased expression of CD44 molecules, however, has been noted in several cancers [16, 19]. Recently, it has been shown that the standard CD44 (CD44s)and variants (v) 4, 6, and 9 were expressed in normal cervical squamous epithelium and in precancerous lesions, but there was a decreased expression of CD44s, v4, and v6 and an increased expression of v9 in precancerous lesions [30]. The expressed variants 6, 7, and 8 were reported to be associated with poorer prognosis in invasive cervical cancer [31–34]. Studies on the expression of standard CD44 molecule and variant isoforms either in normal or in neoplastic endocervical epithelium, however, are still lacking. The current study was designed to examine the immunoreactivity of CD44 standard and several CD44 splice variant molecules, v4, v6, v7, and v9, in invasive (n 5 22) and in situ (n 5 17) adenocarcinomas of the uterine endocervix. MATERIALS AND METHODS Tissue and Patient Demographics Endocervical tissue samples from 39 patients treated at the University of Kansas Medical Center from April 1986 to March 1997 were retrieved from the archives of the Department of Pathology and Laboratory Medicine. There were 17 cases exhibiting the characteristic histologic features of adenocarcinoma in situ (AIS) and 22 invasive adenocarcinoma cases. Of the invasive adenocarcinoma cases, 7 were well differentiated, 10 were moderately differentiated, and 5 were poorly differentiated. The histologic types of these tumors included 14 of the common endocervical type, 4 endometrioid, 2 villoglandular, 1 signet ring type, and 1 adenosquamous carcinoma. In the AIS group, 8 patients also had squamous cervical lesions in the same specimens, including 2 moderate dysplasia, 1 severe dysplasia, 1 severe dysplasia with human papillomavirus changes, 2 squamous cell carcinoma in situ, and 2 invasive squamous cell carcinoma. Two patients in the invasive adenocarcinoma (adenoCa) group also had invasive squamous cell carcinoma in the same specimens. The mean ages of the AIS and invasive adenocarcinoma patients were 37 years (range 28 –53) and 53.5 years (range 28 – 86), respectively. Clinically, in the AIS group, 4 patients presented with vaginal spotting, 3 with postcoital bleeding, 1 with dysmenorrhea, and 1 with constipation and 5 were asymptomatic. In the invasive carcinoma group, 12 patients presented with vaginal spotting, 4 with vaginal discharge, and 2 with pelvic pain and 5 were asymptomatic. Sexually transmitted disease history of human papillomavirus and chlamydial infection was documented in the medical records in a small number of patients in both groups. Abnormal cytologic findings were
85
reported in 8 patients with AIS and 6 patients with invasive carcinoma. In the AIS group, treatment included cone biopsy (n 5 7), radical hysterectomy (n 5 4), simple hysterectomy (n 5 5), and radiation therapy (n 5 2). Of the invasive carcinoma group, treatments included radical hysterectomy (n 5 9), simple hysterectomy (n 5 2), radiation therapy alone (n 5 9), chemotherapy (n 5 5), cesium implants (n 5 5), and staging laparotomy with local lymph node dissection plus radiation therapy (n 5 4). Nine women with AIS showed no evidence of disease after treatment for a follow-up period of 3 to 9 years. None of those patients progressed to an invasive lesion. The rest of the AIS cases were not available for followup. Of the invasive carcinoma group, 7 patients died of the disease, 7 were alive without evidence of disease for a follow-up period of 1 to 9 years, and the rest were not available for follow-up. All tissues were fixed in 10% neutral-buffered formalin, embedded in paraffin, and processed to produce 5-mm sections. Immunoperoxidase Staining Monoclonal antibodies against CD44s and CD44 v7 were purchased from DAKO (Carpenteria, CA; 1:100 dilution) and BioSource International (Burlingame, CA; 1:100 dilution). Three hybridoma cell lines (produced by fusing Sp2/0-Ag14 myeloma cells with splenocytes from BALB/c mice) were obtained from the American Type Culture Collection (Rockville, MD). These cell lines (HB-256, HB-257, and HB-258) secrete monoclonal antibodies against the gene products of v6, v4, and v9 exons of human CD44, respectively. The 1-ml supply of cells was thawed and added to 1 ml of growth medium (see below). After centrifugation at 1500 rpm for 5 min the cell pellets were resuspended and grown at recommended dilutions in 20 ml of RPMI (Meditech) with 10% fetal calf serum, 4 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin, 0.25 mg/ml amphotericin B, and sodium pyruvate. After 4 days’ growth, cell cultures were split into three and grown 4 more days. Centrifugation yielded cell-free supernatant. The total protein content of the supernatant was determined by optical densitometry at 280 nm and the proportion thereof representing IgG was quantified by electrophoresis. Supernatants were preserved with 15 mM sodium azide at 4°C and the undiluted supernatants were used in immunohistochemistry. Tonsillar squamous epithelium was used as a positive control, and normal mouse serum without primary antisera (DAKO; undiluted) was used as a negative control. Paraffin sections were air-dried for 30 min, oven-dried, deparaffinized, and rehydrated. Slides were microwaved on high power with ChemMate antigen retrieval solution (BioTek, Santa Barbara, CA). Antibody supernatant was used at full strength or diluted in antibody dilution buffer (BioTek). Immunohistochemical staining was carried out on a Tech Mate 1000 (BioTek) apparatus with ChemMate reagents. Briefly, the sections were deparaffinized in xylene and re-
86
LU ET AL.
FIG. 1. Comparison of the immunostaining pattern of CD44 standard molecule in normal squamous epithelium of the cervix uteri (A), normal endocervix (B), adenocarcinoma in situ (C), and invasive adenocarcinoma of the cervix (D). Normal squamous and endocervical epithelium showed a CD44s staining pattern that was localized to only the basal side of the epithelium (A and B). In contrast, in AIS (C) and invasive adenocarcinoma (D) it was diffusely expressed along the entire cytoplasmic membrane, including the luminal surface. C also shows transition from normal endocervical epithelium in the middle portion of the picture (weak basal staining) to AIS (strong staining in the full layer of epithelium).
hydrated in alcohol, followed by blockage of endogenous peroxidase by incubation with 0.3% hydrogen peroxide for 30 min. The primary antibody was applied to sections for 30 min, followed by incubation with biotinylated anti-immunoglobulin. Enzymatic treatment with 0.04% pepsin (10 min, at 37°C) and 0.1% proteinase K (10 min at 37°C) was performed on the slides stained for CD44s and variants, respectively. For color development, diaminobenzidine hydrogen peroxide was used, creating a brown reaction product. Hematoxylin was used as a counterstain. Quantitation of Staining Intensity Slides were stained with ethyl green for 15 min, cleared in butyl alcohol, rinsed in xylene, and coverslipped. Ten fields were examined on each slide and a predominant staining intensity was used for scoring. Cytoplasmic staining was semiquantitatively scored 0 (absent), 11 (sporadic or weak reaction of single cells), 21 (moderate and heterogeneous staining pattern with less than 50% of positive cells), or 31 (strong
staining of more than 50% of the cells) for a given cell type. The percentages of positive cells were also evaluated and compared among different groups. The statistical difference in the expression of CD44 molecules among different study groups was examined using the Fisher exact test. RESULTS Sections of the cervical specimens from 39 patients contained 17 in situ adenocarcinomas and 22 invasive adenocarcinomas. Normal squamous and endocervical epithelium was present in a majority of the specimens and served as internal controls for immunoreactivity of CD44 molecules. The cervical specimens from 3 nonneoplastic patients served as normal controls. When the control (nonimmunized) normal mouse serum was substituted for mouse monoclonal anti-CD44, no staining was detected from any specimen. There was a quantitative and qualitative difference of CD44s expression between normal and neoplastic cervical epithelium.
87
ALTERED EXPRESSION OF CD44 IN ADENOCARCINOMA
FIG. 2. Immunostaining pattern of CD44 standard molecule in invasive endocervical adenocarcinoma. Weak to moderate staining intensity was noted in low-grade tumors (A and B) and moderate to strong staining was evident in moderate and poorly differentiated tumors (C and D).
Normal squamous epithelium from control, AIS, and invasive adenoCa patients showed a consistent strong staining reaction to CD44s antibody that was limited to the stratum basale and stratum spinosum only (Fig. 1A). Normal endocervical epithelium present in neoplastic and control specimens either lacked immunoreactivity for CD44 standard molecule or showed only a weak staining in the basal portion of the epithelium along the basement membrane, whereas the luminal side remained unstained (Fig. 1B). In contrast, 94% of in situ adenocarcinomas and 95% of invasive adenocarcinomas displayed various levels of expression of CD44 standard molecule. The expression of CD44s has lost its polarity and diffusely involved the entire cytoplasmic membrane, especially along the luminal portion of the neoplastic epithelium (Figs. 1C, 1D, and 2). In some specimens, a distinct transition from normal endocervical epithelium to adenocarcinoma in situ was observed and the expression of CD44s clearly reflected such a transition (Fig. 1C). The proportions of in situ adenocarcinomas and invasive adenocarcinomas that expressed moderate or strong CD44 were statistically different (P , 0.05). Eighty-two percent (18 of 22) of invasive adenocarcinomas exhibited a moderate (21) to strong (31) CD44s expression, whereas only 53% (9 of 17)
of in situ adenocarcinomas showed such expression (Table 1). Quantitatively, 45.7% cells in invasive adenocarcinomas and 22.6% cells in in situ adenocarcinomas expressed CD44s (P , 0.05). Furthermore, albeit not statistically significant, invasive tumors with a higher histologic grade contained more neoplastic cells staining for CD44s and were stained at a higher intensity than those of lower grade and more differentiated tumors (Fig. 2). No significant correlation was found between TABLE 1 Expression of CD44s in in Situ and Invasive Adenocarcinoma of the Endocervix Number (%) of tumors that express CD44s Staining intensity
Adenocarcinoma in situ
Invasive adenocarcinoma
Negative Positive Focal/11 21 31 Total
1 (5.9)
1 (4.6)
7 (41.2) 7 (41.2) 2 (11.7) 17 (100)
3 (13.6) 13 (59.1) 5 (22.7) 22 (100)
88
LU ET AL.
TABLE 2 Expression of CD44v4 in in Situ and Invasive Adenocarcinoma of the Endocervix
TABLE 3 Expression of CD44v6 in in Situ and Invasive Adenocarcinoma of the Endocervix
Number (%) of tumors that express CD44v4
Number (%) of tumors that express CD44v6
Staining intensity
Adenocarcinoma in situ
Invasive adenocarcinoma
Staining intensity
Adenocarcinoma in situ
Invasive adenocarcinoma
Negative Focal positive Glandular Squamous Total
15 (93.8)
17 (77.3)
16 (94.1)
16 (72.7)
0 1 (6.2) 16 (100)
2 (9.1) 3 (13.6) 22 (100)
Negative Positive Focal/11 21 31 Total
1 (5.9) 0 0 17 (100)
3 (13.6) 2 (9.1) 1 (4.6) 22 (100)
the levels of CD44s expression and the metastatic potential of the invasive tumors or patients’ survival (data not shown). In both AIS and invasive adenoCa groups, the levels of CD44s expression were not associated with squamous cervical lesions or human papillomavirus changes (data not shown). Normal endocervical epithelium in all samples exhibited no expression of CD44 variants 4, 6, and 7, but showed a weak expression of CD44v9 in the basal layer. The staining pattern for the isoforms in the normal cervical squamous epithelium was identical to that of the standard molecule. CD44v7 was not detected in any in situ or invasive adenocarcinomas. CD44v4 was focally expressed in 1 of 16 (6%) in situ lesions and 5 of 22 (23%) invasive cancers, mainly in the squamous areas (Table 2). Only 1 of 17 (6%) in situ lesions expressed rare CD44v6 antigen, whereas 6 of 22 (27%) invasive cancers displayed a focal weak to diffuse strong expression of CD44v6 (Table 3). The isoform v9 showed an increased expression in most carcinomas compared to the normal epithelium. Eight of 16 (50%) in situ carcinomas and 15 of 22 (68%) invasive carcinomas showed a moderate to strong expression of CD44v9, but no significant difference in the level of expression was observed between these two groups (Table 4). DISCUSSION Endocervical adenocarcinoma is a disease that is on the rise. It has recently been reported that its relative frequency has tripled in the past 2 decades [1, 3, 35]. Although routine cytologic examination has played a major role in the improvement of our diagnostic skills in the study of endocervical adenocarcinoma and its precursors, the sensitivity of lesional detection remains low in comparison to squamous lesions. The cytologic diagnosis of glandular lesions is particularly problematic. High rate of false negative Pap smears appears to be one of the major concerns in the current system of screening. Furthermore, the inability to separate reactive glandular atypia from atypia due to neoplastic process has been well recognized [36, 37]. Currently, the diagnosis of this tumor relies primarily on morphologic findings with no appropriate potential biomarkers available for the study of tumor progression and invasive-
ness. Invasive adenocarcinoma appears to arise through a multistep process in which endocervical dysplasia and adenocarcinoma in situ are the precursor lesions. In the course of tumor progression altered CD44 protein and a variety of its variant isoforms have been noted in many human tumors, such as skin, urinary bladder, breast, prostate, colon, esophagus, cervix, and others. Standard CD44 protein and/or certain splice variants may become up-regulated, remain unchanged, or even become down-regulated in different tumors. In this context, the study of CD44 and its variant isoforms could serve both as diagnostic and as prognostic indicators of endocervical adenocarcinoma during progression. CD44 standard molecule is expressed by a variety of normal human epithelial and mesenchymal cells such as epidermis [24], squamous mucosa of the uterine cervix [30], and the basal layer of the gastrointestinal epithelium [14, 38]. However, no studies have been done on either normal or neoplastic endocervical epithelium. In the current study, we demonstrated by immunohistochemistry that the standard CD44 molecule and variant isoform 9 are expressed in normal endocervical and squamous epithelium as well as in neoplastic lesions. The expression of the standard form of CD44 and v9 in the normal endocervical epithelium is at a very low level and is limited to only the basal side of the cytoplasm, along the basement membrane. Normal endocervical epithelium exhibited no ex-
TABLE 4 Expression of CD44v9 in in Situ and Invasive Adenocarcinoma of the Endocervix Number (%) of tumors that express CD44v9 Staining intensity
Adenocarcinoma in situ
Invasive adenocarcinoma
Negative Positive Focal/11 21 31 Total
0 (0%)
1 (4.6)
8 (50.0) 5 (31.3) 3 (18.7) 16 (100)
6 (27.2) 7 (31.8) 8 (36.4) 22 (100)
ALTERED EXPRESSION OF CD44 IN ADENOCARCINOMA
pression for variants 4, 6, or 7. These results suggest that CD44s and v9 might play an important role in normal cell– cell and cell–matrix interaction of the endocervix. In contrast to normal endocervical epithelium, immunostaining for standard CD44 and v9 was significantly increased in neoplastic endocervical cells. In addition, invasive adenocarcinomas demonstrated a statistically stronger expression than in situ lesions. CD44v4 and v6 are expressed in a small proportion of invasive adenocarcinomas. A minor aspect of our study was to determine whether a correlation existed between CD44 expression and tumor differentiation. Our results show that poorly differentiated invasive endocervical adenocarcinomas showed a higher CD44s staining intensity than most well-differentiated tumors. There was no significant correlation between the levels of CD44 expression and the metastatic potential of the invasive tumors or patients’ survival (results not shown). These findings indicate an increased expression of CD44 standard molecule associated with a progression through a continuum of morphologic abnormalities beginning with endocervical dysplasia/adenocarcinoma in situ culminating and in the classic invasive adenocarcinoma. Most in situ and invasive adenocarcinomas in our study displayed a luminal as well as a basal expression of CD44 standard molecule. A similar phenomenon was also observed in malignancies from other anatomic locations such as squamous mucosa of uterine cervix [30] and gastrointestinal mucosa [14, 38]. The significance of this altered pattern of expression is not known, but it implicates a special cell– cell or cell–matrix interaction in tumor cells different from those in normal cells. Increased expression of CD44 splice variants has been linked to tumor progression and metastatic potential in a number of different human cancers. For example, increasing incidence of CD44v7/8 epitope expression was observed in human cervical cancers [32–34] and human mammary carcinomas [29], but not in human ovarian cancers [39]. From normal mucosa to colorectal cancers there was an increased expression of CD44v9 and a de novo expression of CD44v6 [13]. CD44v9 was also significantly overexpressed in uterine cervical precancerous lesions [30]. CD44v6 and v9 are expressed in hepatocellular carcinomas and cholangiocarcinomas, but not in normal liver tissues [40]. In summary, our results show for the first time that the expression of CD44s and v9 is frequent in endocervical adenocarcinoma. The different patterns of expression that we observed in neoplastic cells compared to that of normal endocervical glands suggest that CD44 standard molecule and v9 could be involved in the carcinogenesis of endocervical adenocarcinoma. The differential expressions of CD44 and v9 isoform in adenocarcinoma may be used as additional markers for delineating subsets of tumors with more aggressive biology. Similarly, their expression could be used in the differentiation between neoplastic and nonneoplastic glands.
89
REFERENCES 1. Vesterinen E, Forss M, Nieminen U: Increase of cervical adenocarcinoma. A report of 520 cases of cervical carcinomas including 112 tumors with glandular elements. Gynecol Oncol 33:49 –53, 1989 2. Goodman HM, Buttlar CA, Niloff JM, Welch WR, Marck A, Feuer EJ, Lahman EA, Jenison EL, Knapp RC: Adenocarcinoma of the uterine cervix: prognostic factors and patterns of recurrence. Gynecol Oncol 33:241–247, 1989 3. Davis JR, Moon LB: Increased incidence of adenocarcinoma of the uterine cervix. Obstet Gynecol 45:79 – 83, 1975 4. Greer BE, Figge DC, Tamimi HK, Cain JM: Stage 1B adenocarcinoma of the cervix treated by radical hysterectomy and pelvic lymph node dissection. Am J Obstet Gynecol 160:1508 –1514, 1989 5. Reagan JW, Ng ABP: Cellular Detection of Glandular Neoplasms of the Uterine Cervix, in Weid GL (ed): Compendium on Diagnostic Cytology, 6th ed. Chicago, Tutorials of Clinical Cytology, 1988, pp 146 –153 6. Young RH, Scully RE: Invasive adenocarcinoma and related tumors of the uterine cervix. Semin Diagn Pathol 7:205–227, 1990 7. Hurt WG, Silverberg SG, Frable WJ, Belgrad R, Crooks LD Jr: Adenocarcinoma of the cervix. Histologic and clinical features. Am J Obstet Gynecol 129:304 –315, 1977 8. Lesley J, Hyman R, Kincade PW: CD44 and its interaction with extracellular matrix. Adv Immunol 54:271–335, 1993 9. Gunthert U: CD44: a multitude of isoforms with diverse functions. Curr Top Microbiol Immunol 184:47– 63, 1993 10. Haynes BF, Telen MJ, Hale LP, Denning SM: CD44: a molecule involved in leukocyte adherence and T-cell activation. Immunol Today 10:423– 428, 1989 11. Ponta H, Sleeman J, Dall P, Moll J, Sherman L, Herrlich P: CD44 isoforms in metastatic cancer. Invasion Metastasis 14:82– 86, 1995 12. Goodison S, Tarin D: Current status of CD44 variant isoforms as cancer diagnostic markers. Histopathology 32:1– 6, 1998 13. Tanabe KK, Ellis LM, Saya H: Expression of CD44R1 adhesion molecule in colon carcinomas and metastases. Lancet 341:725–726, 1993 14. Kawahara K, Yoshino T, Kawasaki N, Miyake K, Akagi T: Abnormal expression of human CD44 gene in early colorectal malignancy with special reference to variant exon 9 (9v). J Clin Pathol 49:478 – 481, 1996 15. Liu AY: Expression of CD44 in prostate cancer cells. Cancer Lett 76:63– 69, 1994 16. Nagabhushan M, Pretlow TG, Guo Y-J, Amini SB, Pretlow TP: Altered expression of CD44 in human prostate cancer during progression. Am J Clin Pathol 106:647– 651, 1996 17. Zhang XH, Sakamoto H, Takenaka I: Accumulation of p53 and expression of CD44 in human prostatic cancer and benign prostatic hyperplasia: an immunohistochemical study. Br J Urol 77:441– 444, 1996 18. Kan M, Furukawa A, Aki M, Kanayama H, Kagawa S: Expression of CD44 splice variants in bladder cancer. Int J Urol 2:295–301, 1995 19. Sugino T, Gorham H, Yoshida K, Bolodeoku J, Nargund V, Cranston D, Goodison S, Tarin D: Progressive loss of CD44 gene expression in invasive bladder cancer. Am J Pathol 149:873– 882, 1996 20. Heider KH, Ratschek M, Zatloukal K, Adolf GR: Expression of CD44 isoforms in human renal cell carcinomas. Virchows Arch 428:267–273, 1996 21. Terpe HJ, Storkel S, Zimmer U, Anquez V, Fischer C, Pantel K, Gunthert U: Expression of CD44 isoforms in renal cell tumors. Positive correlation to tumor differentiation. Am J Pathol 148:453– 463, 1996 22. Kaaijk P, Troost D, Morsink F, Keehnen RM, Leenstra S, Bosch DA, Pals ST: Expression of CD44 splice variants in human primary brain tumors. J Neurooncol 26:185–190, 1995
90
LU ET AL.
23. Eibl RH, Pietsch T, Moll J, Skroch Angel P, Heider KH, von Ammon K, Wiestler OD, Ponta H, Kleihues P, Herrlich P: Expression of variant CD44 epitopes in human astrocytic brain tumors. J Neurooncol 26:165–170, 1995 24. Seelentag WK, Gunthert U, Saremaslani P, Futo E, Pfaltz M, Heitz PU, Roth J: CD44 standard and variant isoform expression in human epidermal skin tumors is not correlated with tumor aggressiveness but down-regulated during proliferation and tumor de-differentiation. Int J Cancer 69: 218 –224, 1996 25. Prieto VG, Reed JA, McNutt NS, Bogdany JK, Lugo J, Shea CR: Differential expression of CD44 in malignant cutaneous epithelial neoplasms. Am J Dermatopathol 17:447– 451, 1995 26. Korabiowska M, Brinck U, Honig J, Mora O, Bartkowski S, Schauer A: CD-44: a marker of metastases of head and neck melanomas. In Vivo 9:253–256, 1995 27. Perin T, Canzonieri V, Massarut S, Bidoli E, Rossi C, Roncadin M, Carbone A: Immunohistochemical evaluation of multiple biological markers in ductal carcinoma in situ of the breast. Eur J Cancer 32A:1148 –1155, 1996 28. de la Torre M, Heldin P, Bergh J: Expression of the CD44 glycoprotein (lymphocyte-homing receptor) in untreated human breast cancer and its relationship to prognostic markers. Anticancer Res 15:2791–2795, 1995 29. Sinn HP, Heider KH, Skroch-Angel P, von Minckwitz G, Kaufmann M, Herrlich P, Ponta H: Human mammary carcinomas express homologues of rat metastasis-associated variants of CD44. Breast Cancer Res Treat 36:307–313, 1995 30. Dellas A, Schultheiss E, Almendral AC, Torhost J, Gudat F: Expression of CD44 and variant isoforms in cervical intraepithelial neoplasia. Gynecol Oncol 62:281–255, 1996 31. Dall P, Heider KH, Hekele A, Minckwitz G, Kaufmann M, Ponta H, Herrlich P: Surface protein expression and messenger RNA-splicing anal-
ysis of CD44 in uterine cervical cancer and normal cervical epithelium. Cancer Res 54:3337–3341, 1994 32. Dall P, Hekele A, Ikenberg H, Goppinger A, Bauknecht T, Pfleiderer A, Moll J, Hofmann M, Ponta H, Herrlich P: Increasing incidence of CD44v7/8 epitope expression during uterine cervical carcinogenesis. Int J Cancer 69:79 – 85, 1996 33. Kainz C, Kohlberger P, Sliutz G, Tempfer C, Heinzl H, Reinthaller A, Breitenecker G, Koelbl H: Splice variants of CD44 in human cervical cancer stage IB to IIB. Gynecol Oncol 57:382–387, 1995 34. Kainz C, Kohlberger P, Tempfer C, Sliutz G, Gitsch G, Reinthaller A, Breitenecker G: Prognostic value of CD44 splice variants in human stage III cervical cancer. Eur J Cancer 31A:1706 –1709, 1995 35. McGonigle KF, Berek JS: Early stage squamous cell and adenocarcinoma of the cervix. Curr Opin Obstet Gynecol 4:109 –119, 1992 36. Kennedy AW, Salmieri SS, Wirth SL, Biscotti CV, Tuason LJ, Travarca MJ: Results of the clinical evaluation of atypical glandular cells of undetermined significance (AGCUS) detected on cervical cytology screening. Gynecol Oncol 63:14 –18, 1996 37. Lee KR, Manna EA, John T St: Atypical endocervical glandular cells: accuracy of cytologic diagnosis. Diagn Cytopathol 13:202–208, 1995 38. Castella E, Ariza A, Fernandez-Vasalo A, Roca X, Ojanguren I: Expression of CD44H and CD44v3 in normal oesophagus, Barrett mucosa and oesophageal carcinoma. J Clin Pathol 49:489 – 492, 1996 39. Sliutz G, Tempfer C, Winkler S, Kohlberger P, Reinthaller A, Kainz C: Immunohistochemical and serological evaluation of CD44 splice variants in human ovarian cancer. Br J Cancer 72:1494 –1497, 1995 40. Seelentag WK, Flury R, Schmid M, Komminoth P, Saremaslani P, Gunthert U, Heitz PU, Roth J: Expression of CD44 isoforms in normal human liver and also in regenerative and neoplastic liver changes. Verh Dtsch Ges Pathol 79:144 –147, 1995
Altered Expression of CD44 and Variant Isoforms in Human Adenocarcinoma of the Endocervix during Progression 1 Di Lu, M.D., Ph.D.,* Ossama Tawfik, M.D., Ph.D.,* ,† Cooley Pantazis, M.D.,‡ Wendy Hobart, M.D.,† Julia Chapman, M.D.,† and Kenneth Iczkowski, M.D.§ *Department of Pathology and Laboratory Medicine and †Department of Obstetrics and Gynecology, University of Kansas Medical Center, Kansas City, Kansas 66160; ‡Department of Pathology, University of Florida, Gainesville, Florida 32601, and Department of Pathology, Munroe Regional Medical Center, Ocala, Florida 34478; and §Department of Pathology, Reading Hospital and Medical Center, Reading, Pennsylvania 19612 Received November 18, 1998
INTRODUCTION
Objective. Altered expression of the CD44 family of cell adhesion molecules has been associated with tumor progression and metastasis. The aim of this study was to investigate the expression of the gene products of CD44 standard (CD44s) and several alternatively spliced variants (CD44v4, v6, v7, and v9) in adenocarcinoma of the endocervix and to correlate the degree of their expression with disease progression. Methods. Immunohistochemical staining for CD44s and CD44v4, v6, v7, and v9 was performed on formalin-fixed, paraffin-embedded endocervical specimens. Seventeen cases of adenocarcinoma in situ (AIS) and 22 cases of invasive adenocarcinoma of the endocervix were included in this study, and the immunoreactivity was compared with that of normal endocervical epithelium. Results. (1) In the normal endocervical mucosa, immunoreactivity for CD44s and the splice variants was lacking or was confined to only the basal portion of the glandular epithelium along the basement membrane; (2) CD44s was diffusely expressed along the entire cytoplasmic membrane, including the luminal surface of the tumorous glands in 94% of AIS and 95% of invasive adenocarcinomas; (3) a significantly stronger expression of CD44s was observed in invasive adenocarcinomas than in AIS; (4) in contrast to all other splice variants, CD44v9 demonstrated an increased expression in nearly all in situ and invasive lesions compared to the normal tissue; (5) CD44v4 and v6 were expressed only in a small proportion of invasive adenocarcinomas and were near totally absent in the in situ lesions; and (6) CD44 v7 was totally absent in all normal, in situ, and invasive lesions studied. Conclusions. It appears that neoplastic transformation of endocervical epithelium is associated with qualitative and quantitative changes in the expression of CD44 standard molecule and some CD44 splice variants. © 1999 Academic Press Key Words: CD44; variant isoforms 4, 6, 7, and 9; endocervical adenocarcinoma.
Adenocarcinoma of the endocervix is a relatively common gynecologic malignancy that constitutes 7 to 26% of primary cervical carcinomas [1–5]. As is well known for squamous epithelium, it appears that endocervical epithelium progresses through a continuum of morphologic abnormalities beginning with endocervical dysplasia/adenocarcinoma in situ and culminating in the classic invasive adenocarcinoma. Clinically, the majority of preinvasive lesions and up to 20% of invasive adenocarcinomas are asymptomatic and are usually invisible on routine colposcopic examination [6, 7]. Otherwise, patients present with symptoms such as abnormal uterine bleeding, vaginal discharge, or pain. Epidemiologically, endocervical lesions have shown a definite age progression with increasing lesional severity. Cervical glandular lesions, particularly adenocarcinoma in situ, are associated with squamous intraepithelial lesion and human papillomavirus infection. However, the pathogenesis of endocervical adenocarcinoma is not well understood. Tumor development is usually associated with dysfunctional expression or mutation of various genes, including oncogene overexpression, aberrant tumor suppressor gene expression, or deletion and the inappropriate expression of a variety of cytokines and growth factors. CD44 is a family of transmembrane glycoproteins involved in cell– cell and cell– extracellular matrix interaction [8]. The CD44 gene, which maps to chromosome 11, contains at least 21 exons spanning about 60 kb and can be subdivided into five structural domains. The standard CD44 molecule plays an important role in lymphocyte homing, tissue regeneration, and tumor cell dissemination [8 –12]. At least 25 different CD44 variant isoforms can be derived from alternative splicing and more heterogeneity results from extensive posttranslational modification of the protein products. The normal functions of CD44 variants are entirely unknown, however. Altered expression of both standard molecule and certain splice variants of CD44 has been linked with invasive and metastatic behaviors of several human malignancies such as colorectal carcinoma
1 Part of this article was presented in poster form at the Annual Fall Meeting of the American Society of Clinical Pathologists, Philadelphia, September 1997.
0090-8258/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
84
ALTERED EXPRESSION OF CD44 IN ADENOCARCINOMA
[13, 14], prostatic adenocarcinoma [15–17], bladder cancer [18, 19], renal cell carcinoma [20, 21], brain tumors [22, 23], skin cancers [24 –26], and breast cancer [27–29]. In most studies, increased expression of CD44 standard and/or variants correlated with progression and metastatic behavior of malignancies. Decreased expression of CD44 molecules, however, has been noted in several cancers [16, 19]. Recently, it has been shown that the standard CD44 (CD44s)and variants (v) 4, 6, and 9 were expressed in normal cervical squamous epithelium and in precancerous lesions, but there was a decreased expression of CD44s, v4, and v6 and an increased expression of v9 in precancerous lesions [30]. The expressed variants 6, 7, and 8 were reported to be associated with poorer prognosis in invasive cervical cancer [31–34]. Studies on the expression of standard CD44 molecule and variant isoforms either in normal or in neoplastic endocervical epithelium, however, are still lacking. The current study was designed to examine the immunoreactivity of CD44 standard and several CD44 splice variant molecules, v4, v6, v7, and v9, in invasive (n 5 22) and in situ (n 5 17) adenocarcinomas of the uterine endocervix. MATERIALS AND METHODS Tissue and Patient Demographics Endocervical tissue samples from 39 patients treated at the University of Kansas Medical Center from April 1986 to March 1997 were retrieved from the archives of the Department of Pathology and Laboratory Medicine. There were 17 cases exhibiting the characteristic histologic features of adenocarcinoma in situ (AIS) and 22 invasive adenocarcinoma cases. Of the invasive adenocarcinoma cases, 7 were well differentiated, 10 were moderately differentiated, and 5 were poorly differentiated. The histologic types of these tumors included 14 of the common endocervical type, 4 endometrioid, 2 villoglandular, 1 signet ring type, and 1 adenosquamous carcinoma. In the AIS group, 8 patients also had squamous cervical lesions in the same specimens, including 2 moderate dysplasia, 1 severe dysplasia, 1 severe dysplasia with human papillomavirus changes, 2 squamous cell carcinoma in situ, and 2 invasive squamous cell carcinoma. Two patients in the invasive adenocarcinoma (adenoCa) group also had invasive squamous cell carcinoma in the same specimens. The mean ages of the AIS and invasive adenocarcinoma patients were 37 years (range 28 –53) and 53.5 years (range 28 – 86), respectively. Clinically, in the AIS group, 4 patients presented with vaginal spotting, 3 with postcoital bleeding, 1 with dysmenorrhea, and 1 with constipation and 5 were asymptomatic. In the invasive carcinoma group, 12 patients presented with vaginal spotting, 4 with vaginal discharge, and 2 with pelvic pain and 5 were asymptomatic. Sexually transmitted disease history of human papillomavirus and chlamydial infection was documented in the medical records in a small number of patients in both groups. Abnormal cytologic findings were
85
reported in 8 patients with AIS and 6 patients with invasive carcinoma. In the AIS group, treatment included cone biopsy (n 5 7), radical hysterectomy (n 5 4), simple hysterectomy (n 5 5), and radiation therapy (n 5 2). Of the invasive carcinoma group, treatments included radical hysterectomy (n 5 9), simple hysterectomy (n 5 2), radiation therapy alone (n 5 9), chemotherapy (n 5 5), cesium implants (n 5 5), and staging laparotomy with local lymph node dissection plus radiation therapy (n 5 4). Nine women with AIS showed no evidence of disease after treatment for a follow-up period of 3 to 9 years. None of those patients progressed to an invasive lesion. The rest of the AIS cases were not available for followup. Of the invasive carcinoma group, 7 patients died of the disease, 7 were alive without evidence of disease for a follow-up period of 1 to 9 years, and the rest were not available for follow-up. All tissues were fixed in 10% neutral-buffered formalin, embedded in paraffin, and processed to produce 5-mm sections. Immunoperoxidase Staining Monoclonal antibodies against CD44s and CD44 v7 were purchased from DAKO (Carpenteria, CA; 1:100 dilution) and BioSource International (Burlingame, CA; 1:100 dilution). Three hybridoma cell lines (produced by fusing Sp2/0-Ag14 myeloma cells with splenocytes from BALB/c mice) were obtained from the American Type Culture Collection (Rockville, MD). These cell lines (HB-256, HB-257, and HB-258) secrete monoclonal antibodies against the gene products of v6, v4, and v9 exons of human CD44, respectively. The 1-ml supply of cells was thawed and added to 1 ml of growth medium (see below). After centrifugation at 1500 rpm for 5 min the cell pellets were resuspended and grown at recommended dilutions in 20 ml of RPMI (Meditech) with 10% fetal calf serum, 4 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin, 0.25 mg/ml amphotericin B, and sodium pyruvate. After 4 days’ growth, cell cultures were split into three and grown 4 more days. Centrifugation yielded cell-free supernatant. The total protein content of the supernatant was determined by optical densitometry at 280 nm and the proportion thereof representing IgG was quantified by electrophoresis. Supernatants were preserved with 15 mM sodium azide at 4°C and the undiluted supernatants were used in immunohistochemistry. Tonsillar squamous epithelium was used as a positive control, and normal mouse serum without primary antisera (DAKO; undiluted) was used as a negative control. Paraffin sections were air-dried for 30 min, oven-dried, deparaffinized, and rehydrated. Slides were microwaved on high power with ChemMate antigen retrieval solution (BioTek, Santa Barbara, CA). Antibody supernatant was used at full strength or diluted in antibody dilution buffer (BioTek). Immunohistochemical staining was carried out on a Tech Mate 1000 (BioTek) apparatus with ChemMate reagents. Briefly, the sections were deparaffinized in xylene and re-
86
LU ET AL.
FIG. 1. Comparison of the immunostaining pattern of CD44 standard molecule in normal squamous epithelium of the cervix uteri (A), normal endocervix (B), adenocarcinoma in situ (C), and invasive adenocarcinoma of the cervix (D). Normal squamous and endocervical epithelium showed a CD44s staining pattern that was localized to only the basal side of the epithelium (A and B). In contrast, in AIS (C) and invasive adenocarcinoma (D) it was diffusely expressed along the entire cytoplasmic membrane, including the luminal surface. C also shows transition from normal endocervical epithelium in the middle portion of the picture (weak basal staining) to AIS (strong staining in the full layer of epithelium).
hydrated in alcohol, followed by blockage of endogenous peroxidase by incubation with 0.3% hydrogen peroxide for 30 min. The primary antibody was applied to sections for 30 min, followed by incubation with biotinylated anti-immunoglobulin. Enzymatic treatment with 0.04% pepsin (10 min, at 37°C) and 0.1% proteinase K (10 min at 37°C) was performed on the slides stained for CD44s and variants, respectively. For color development, diaminobenzidine hydrogen peroxide was used, creating a brown reaction product. Hematoxylin was used as a counterstain. Quantitation of Staining Intensity Slides were stained with ethyl green for 15 min, cleared in butyl alcohol, rinsed in xylene, and coverslipped. Ten fields were examined on each slide and a predominant staining intensity was used for scoring. Cytoplasmic staining was semiquantitatively scored 0 (absent), 11 (sporadic or weak reaction of single cells), 21 (moderate and heterogeneous staining pattern with less than 50% of positive cells), or 31 (strong
staining of more than 50% of the cells) for a given cell type. The percentages of positive cells were also evaluated and compared among different groups. The statistical difference in the expression of CD44 molecules among different study groups was examined using the Fisher exact test. RESULTS Sections of the cervical specimens from 39 patients contained 17 in situ adenocarcinomas and 22 invasive adenocarcinomas. Normal squamous and endocervical epithelium was present in a majority of the specimens and served as internal controls for immunoreactivity of CD44 molecules. The cervical specimens from 3 nonneoplastic patients served as normal controls. When the control (nonimmunized) normal mouse serum was substituted for mouse monoclonal anti-CD44, no staining was detected from any specimen. There was a quantitative and qualitative difference of CD44s expression between normal and neoplastic cervical epithelium.
87
ALTERED EXPRESSION OF CD44 IN ADENOCARCINOMA
FIG. 2. Immunostaining pattern of CD44 standard molecule in invasive endocervical adenocarcinoma. Weak to moderate staining intensity was noted in low-grade tumors (A and B) and moderate to strong staining was evident in moderate and poorly differentiated tumors (C and D).
Normal squamous epithelium from control, AIS, and invasive adenoCa patients showed a consistent strong staining reaction to CD44s antibody that was limited to the stratum basale and stratum spinosum only (Fig. 1A). Normal endocervical epithelium present in neoplastic and control specimens either lacked immunoreactivity for CD44 standard molecule or showed only a weak staining in the basal portion of the epithelium along the basement membrane, whereas the luminal side remained unstained (Fig. 1B). In contrast, 94% of in situ adenocarcinomas and 95% of invasive adenocarcinomas displayed various levels of expression of CD44 standard molecule. The expression of CD44s has lost its polarity and diffusely involved the entire cytoplasmic membrane, especially along the luminal portion of the neoplastic epithelium (Figs. 1C, 1D, and 2). In some specimens, a distinct transition from normal endocervical epithelium to adenocarcinoma in situ was observed and the expression of CD44s clearly reflected such a transition (Fig. 1C). The proportions of in situ adenocarcinomas and invasive adenocarcinomas that expressed moderate or strong CD44 were statistically different (P , 0.05). Eighty-two percent (18 of 22) of invasive adenocarcinomas exhibited a moderate (21) to strong (31) CD44s expression, whereas only 53% (9 of 17)
of in situ adenocarcinomas showed such expression (Table 1). Quantitatively, 45.7% cells in invasive adenocarcinomas and 22.6% cells in in situ adenocarcinomas expressed CD44s (P , 0.05). Furthermore, albeit not statistically significant, invasive tumors with a higher histologic grade contained more neoplastic cells staining for CD44s and were stained at a higher intensity than those of lower grade and more differentiated tumors (Fig. 2). No significant correlation was found between TABLE 1 Expression of CD44s in in Situ and Invasive Adenocarcinoma of the Endocervix Number (%) of tumors that express CD44s Staining intensity
Adenocarcinoma in situ
Invasive adenocarcinoma
Negative Positive Focal/11 21 31 Total
1 (5.9)
1 (4.6)
7 (41.2) 7 (41.2) 2 (11.7) 17 (100)
3 (13.6) 13 (59.1) 5 (22.7) 22 (100)
88
LU ET AL.
TABLE 2 Expression of CD44v4 in in Situ and Invasive Adenocarcinoma of the Endocervix
TABLE 3 Expression of CD44v6 in in Situ and Invasive Adenocarcinoma of the Endocervix
Number (%) of tumors that express CD44v4
Number (%) of tumors that express CD44v6
Staining intensity
Adenocarcinoma in situ
Invasive adenocarcinoma
Staining intensity
Adenocarcinoma in situ
Invasive adenocarcinoma
Negative Focal positive Glandular Squamous Total
15 (93.8)
17 (77.3)
16 (94.1)
16 (72.7)
0 1 (6.2) 16 (100)
2 (9.1) 3 (13.6) 22 (100)
Negative Positive Focal/11 21 31 Total
1 (5.9) 0 0 17 (100)
3 (13.6) 2 (9.1) 1 (4.6) 22 (100)
the levels of CD44s expression and the metastatic potential of the invasive tumors or patients’ survival (data not shown). In both AIS and invasive adenoCa groups, the levels of CD44s expression were not associated with squamous cervical lesions or human papillomavirus changes (data not shown). Normal endocervical epithelium in all samples exhibited no expression of CD44 variants 4, 6, and 7, but showed a weak expression of CD44v9 in the basal layer. The staining pattern for the isoforms in the normal cervical squamous epithelium was identical to that of the standard molecule. CD44v7 was not detected in any in situ or invasive adenocarcinomas. CD44v4 was focally expressed in 1 of 16 (6%) in situ lesions and 5 of 22 (23%) invasive cancers, mainly in the squamous areas (Table 2). Only 1 of 17 (6%) in situ lesions expressed rare CD44v6 antigen, whereas 6 of 22 (27%) invasive cancers displayed a focal weak to diffuse strong expression of CD44v6 (Table 3). The isoform v9 showed an increased expression in most carcinomas compared to the normal epithelium. Eight of 16 (50%) in situ carcinomas and 15 of 22 (68%) invasive carcinomas showed a moderate to strong expression of CD44v9, but no significant difference in the level of expression was observed between these two groups (Table 4). DISCUSSION Endocervical adenocarcinoma is a disease that is on the rise. It has recently been reported that its relative frequency has tripled in the past 2 decades [1, 3, 35]. Although routine cytologic examination has played a major role in the improvement of our diagnostic skills in the study of endocervical adenocarcinoma and its precursors, the sensitivity of lesional detection remains low in comparison to squamous lesions. The cytologic diagnosis of glandular lesions is particularly problematic. High rate of false negative Pap smears appears to be one of the major concerns in the current system of screening. Furthermore, the inability to separate reactive glandular atypia from atypia due to neoplastic process has been well recognized [36, 37]. Currently, the diagnosis of this tumor relies primarily on morphologic findings with no appropriate potential biomarkers available for the study of tumor progression and invasive-
ness. Invasive adenocarcinoma appears to arise through a multistep process in which endocervical dysplasia and adenocarcinoma in situ are the precursor lesions. In the course of tumor progression altered CD44 protein and a variety of its variant isoforms have been noted in many human tumors, such as skin, urinary bladder, breast, prostate, colon, esophagus, cervix, and others. Standard CD44 protein and/or certain splice variants may become up-regulated, remain unchanged, or even become down-regulated in different tumors. In this context, the study of CD44 and its variant isoforms could serve both as diagnostic and as prognostic indicators of endocervical adenocarcinoma during progression. CD44 standard molecule is expressed by a variety of normal human epithelial and mesenchymal cells such as epidermis [24], squamous mucosa of the uterine cervix [30], and the basal layer of the gastrointestinal epithelium [14, 38]. However, no studies have been done on either normal or neoplastic endocervical epithelium. In the current study, we demonstrated by immunohistochemistry that the standard CD44 molecule and variant isoform 9 are expressed in normal endocervical and squamous epithelium as well as in neoplastic lesions. The expression of the standard form of CD44 and v9 in the normal endocervical epithelium is at a very low level and is limited to only the basal side of the cytoplasm, along the basement membrane. Normal endocervical epithelium exhibited no ex-
TABLE 4 Expression of CD44v9 in in Situ and Invasive Adenocarcinoma of the Endocervix Number (%) of tumors that express CD44v9 Staining intensity
Adenocarcinoma in situ
Invasive adenocarcinoma
Negative Positive Focal/11 21 31 Total
0 (0%)
1 (4.6)
8 (50.0) 5 (31.3) 3 (18.7) 16 (100)
6 (27.2) 7 (31.8) 8 (36.4) 22 (100)
ALTERED EXPRESSION OF CD44 IN ADENOCARCINOMA
pression for variants 4, 6, or 7. These results suggest that CD44s and v9 might play an important role in normal cell– cell and cell–matrix interaction of the endocervix. In contrast to normal endocervical epithelium, immunostaining for standard CD44 and v9 was significantly increased in neoplastic endocervical cells. In addition, invasive adenocarcinomas demonstrated a statistically stronger expression than in situ lesions. CD44v4 and v6 are expressed in a small proportion of invasive adenocarcinomas. A minor aspect of our study was to determine whether a correlation existed between CD44 expression and tumor differentiation. Our results show that poorly differentiated invasive endocervical adenocarcinomas showed a higher CD44s staining intensity than most well-differentiated tumors. There was no significant correlation between the levels of CD44 expression and the metastatic potential of the invasive tumors or patients’ survival (results not shown). These findings indicate an increased expression of CD44 standard molecule associated with a progression through a continuum of morphologic abnormalities beginning with endocervical dysplasia/adenocarcinoma in situ culminating and in the classic invasive adenocarcinoma. Most in situ and invasive adenocarcinomas in our study displayed a luminal as well as a basal expression of CD44 standard molecule. A similar phenomenon was also observed in malignancies from other anatomic locations such as squamous mucosa of uterine cervix [30] and gastrointestinal mucosa [14, 38]. The significance of this altered pattern of expression is not known, but it implicates a special cell– cell or cell–matrix interaction in tumor cells different from those in normal cells. Increased expression of CD44 splice variants has been linked to tumor progression and metastatic potential in a number of different human cancers. For example, increasing incidence of CD44v7/8 epitope expression was observed in human cervical cancers [32–34] and human mammary carcinomas [29], but not in human ovarian cancers [39]. From normal mucosa to colorectal cancers there was an increased expression of CD44v9 and a de novo expression of CD44v6 [13]. CD44v9 was also significantly overexpressed in uterine cervical precancerous lesions [30]. CD44v6 and v9 are expressed in hepatocellular carcinomas and cholangiocarcinomas, but not in normal liver tissues [40]. In summary, our results show for the first time that the expression of CD44s and v9 is frequent in endocervical adenocarcinoma. The different patterns of expression that we observed in neoplastic cells compared to that of normal endocervical glands suggest that CD44 standard molecule and v9 could be involved in the carcinogenesis of endocervical adenocarcinoma. The differential expressions of CD44 and v9 isoform in adenocarcinoma may be used as additional markers for delineating subsets of tumors with more aggressive biology. Similarly, their expression could be used in the differentiation between neoplastic and nonneoplastic glands.
89
REFERENCES 1. Vesterinen E, Forss M, Nieminen U: Increase of cervical adenocarcinoma. A report of 520 cases of cervical carcinomas including 112 tumors with glandular elements. Gynecol Oncol 33:49 –53, 1989 2. Goodman HM, Buttlar CA, Niloff JM, Welch WR, Marck A, Feuer EJ, Lahman EA, Jenison EL, Knapp RC: Adenocarcinoma of the uterine cervix: prognostic factors and patterns of recurrence. Gynecol Oncol 33:241–247, 1989 3. Davis JR, Moon LB: Increased incidence of adenocarcinoma of the uterine cervix. Obstet Gynecol 45:79 – 83, 1975 4. Greer BE, Figge DC, Tamimi HK, Cain JM: Stage 1B adenocarcinoma of the cervix treated by radical hysterectomy and pelvic lymph node dissection. Am J Obstet Gynecol 160:1508 –1514, 1989 5. Reagan JW, Ng ABP: Cellular Detection of Glandular Neoplasms of the Uterine Cervix, in Weid GL (ed): Compendium on Diagnostic Cytology, 6th ed. Chicago, Tutorials of Clinical Cytology, 1988, pp 146 –153 6. Young RH, Scully RE: Invasive adenocarcinoma and related tumors of the uterine cervix. Semin Diagn Pathol 7:205–227, 1990 7. Hurt WG, Silverberg SG, Frable WJ, Belgrad R, Crooks LD Jr: Adenocarcinoma of the cervix. Histologic and clinical features. Am J Obstet Gynecol 129:304 –315, 1977 8. Lesley J, Hyman R, Kincade PW: CD44 and its interaction with extracellular matrix. Adv Immunol 54:271–335, 1993 9. Gunthert U: CD44: a multitude of isoforms with diverse functions. Curr Top Microbiol Immunol 184:47– 63, 1993 10. Haynes BF, Telen MJ, Hale LP, Denning SM: CD44: a molecule involved in leukocyte adherence and T-cell activation. Immunol Today 10:423– 428, 1989 11. Ponta H, Sleeman J, Dall P, Moll J, Sherman L, Herrlich P: CD44 isoforms in metastatic cancer. Invasion Metastasis 14:82– 86, 1995 12. Goodison S, Tarin D: Current status of CD44 variant isoforms as cancer diagnostic markers. Histopathology 32:1– 6, 1998 13. Tanabe KK, Ellis LM, Saya H: Expression of CD44R1 adhesion molecule in colon carcinomas and metastases. Lancet 341:725–726, 1993 14. Kawahara K, Yoshino T, Kawasaki N, Miyake K, Akagi T: Abnormal expression of human CD44 gene in early colorectal malignancy with special reference to variant exon 9 (9v). J Clin Pathol 49:478 – 481, 1996 15. Liu AY: Expression of CD44 in prostate cancer cells. Cancer Lett 76:63– 69, 1994 16. Nagabhushan M, Pretlow TG, Guo Y-J, Amini SB, Pretlow TP: Altered expression of CD44 in human prostate cancer during progression. Am J Clin Pathol 106:647– 651, 1996 17. Zhang XH, Sakamoto H, Takenaka I: Accumulation of p53 and expression of CD44 in human prostatic cancer and benign prostatic hyperplasia: an immunohistochemical study. Br J Urol 77:441– 444, 1996 18. Kan M, Furukawa A, Aki M, Kanayama H, Kagawa S: Expression of CD44 splice variants in bladder cancer. Int J Urol 2:295–301, 1995 19. Sugino T, Gorham H, Yoshida K, Bolodeoku J, Nargund V, Cranston D, Goodison S, Tarin D: Progressive loss of CD44 gene expression in invasive bladder cancer. Am J Pathol 149:873– 882, 1996 20. Heider KH, Ratschek M, Zatloukal K, Adolf GR: Expression of CD44 isoforms in human renal cell carcinomas. Virchows Arch 428:267–273, 1996 21. Terpe HJ, Storkel S, Zimmer U, Anquez V, Fischer C, Pantel K, Gunthert U: Expression of CD44 isoforms in renal cell tumors. Positive correlation to tumor differentiation. Am J Pathol 148:453– 463, 1996 22. Kaaijk P, Troost D, Morsink F, Keehnen RM, Leenstra S, Bosch DA, Pals ST: Expression of CD44 splice variants in human primary brain tumors. J Neurooncol 26:185–190, 1995
90
LU ET AL.
23. Eibl RH, Pietsch T, Moll J, Skroch Angel P, Heider KH, von Ammon K, Wiestler OD, Ponta H, Kleihues P, Herrlich P: Expression of variant CD44 epitopes in human astrocytic brain tumors. J Neurooncol 26:165–170, 1995 24. Seelentag WK, Gunthert U, Saremaslani P, Futo E, Pfaltz M, Heitz PU, Roth J: CD44 standard and variant isoform expression in human epidermal skin tumors is not correlated with tumor aggressiveness but down-regulated during proliferation and tumor de-differentiation. Int J Cancer 69: 218 –224, 1996 25. Prieto VG, Reed JA, McNutt NS, Bogdany JK, Lugo J, Shea CR: Differential expression of CD44 in malignant cutaneous epithelial neoplasms. Am J Dermatopathol 17:447– 451, 1995 26. Korabiowska M, Brinck U, Honig J, Mora O, Bartkowski S, Schauer A: CD-44: a marker of metastases of head and neck melanomas. In Vivo 9:253–256, 1995 27. Perin T, Canzonieri V, Massarut S, Bidoli E, Rossi C, Roncadin M, Carbone A: Immunohistochemical evaluation of multiple biological markers in ductal carcinoma in situ of the breast. Eur J Cancer 32A:1148 –1155, 1996 28. de la Torre M, Heldin P, Bergh J: Expression of the CD44 glycoprotein (lymphocyte-homing receptor) in untreated human breast cancer and its relationship to prognostic markers. Anticancer Res 15:2791–2795, 1995 29. Sinn HP, Heider KH, Skroch-Angel P, von Minckwitz G, Kaufmann M, Herrlich P, Ponta H: Human mammary carcinomas express homologues of rat metastasis-associated variants of CD44. Breast Cancer Res Treat 36:307–313, 1995 30. Dellas A, Schultheiss E, Almendral AC, Torhost J, Gudat F: Expression of CD44 and variant isoforms in cervical intraepithelial neoplasia. Gynecol Oncol 62:281–255, 1996 31. Dall P, Heider KH, Hekele A, Minckwitz G, Kaufmann M, Ponta H, Herrlich P: Surface protein expression and messenger RNA-splicing anal-
ysis of CD44 in uterine cervical cancer and normal cervical epithelium. Cancer Res 54:3337–3341, 1994 32. Dall P, Hekele A, Ikenberg H, Goppinger A, Bauknecht T, Pfleiderer A, Moll J, Hofmann M, Ponta H, Herrlich P: Increasing incidence of CD44v7/8 epitope expression during uterine cervical carcinogenesis. Int J Cancer 69:79 – 85, 1996 33. Kainz C, Kohlberger P, Sliutz G, Tempfer C, Heinzl H, Reinthaller A, Breitenecker G, Koelbl H: Splice variants of CD44 in human cervical cancer stage IB to IIB. Gynecol Oncol 57:382–387, 1995 34. Kainz C, Kohlberger P, Tempfer C, Sliutz G, Gitsch G, Reinthaller A, Breitenecker G: Prognostic value of CD44 splice variants in human stage III cervical cancer. Eur J Cancer 31A:1706 –1709, 1995 35. McGonigle KF, Berek JS: Early stage squamous cell and adenocarcinoma of the cervix. Curr Opin Obstet Gynecol 4:109 –119, 1992 36. Kennedy AW, Salmieri SS, Wirth SL, Biscotti CV, Tuason LJ, Travarca MJ: Results of the clinical evaluation of atypical glandular cells of undetermined significance (AGCUS) detected on cervical cytology screening. Gynecol Oncol 63:14 –18, 1996 37. Lee KR, Manna EA, John T St: Atypical endocervical glandular cells: accuracy of cytologic diagnosis. Diagn Cytopathol 13:202–208, 1995 38. Castella E, Ariza A, Fernandez-Vasalo A, Roca X, Ojanguren I: Expression of CD44H and CD44v3 in normal oesophagus, Barrett mucosa and oesophageal carcinoma. J Clin Pathol 49:489 – 492, 1996 39. Sliutz G, Tempfer C, Winkler S, Kohlberger P, Reinthaller A, Kainz C: Immunohistochemical and serological evaluation of CD44 splice variants in human ovarian cancer. Br J Cancer 72:1494 –1497, 1995 40. Seelentag WK, Flury R, Schmid M, Komminoth P, Saremaslani P, Gunthert U, Heitz PU, Roth J: Expression of CD44 isoforms in normal human liver and also in regenerative and neoplastic liver changes. Verh Dtsch Ges Pathol 79:144 –147, 1995