| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Imaging, Diagnosis, Prognosis |
Authors' Affiliations: Departments of 1 Oncological and Surgical Sciences (Pathology) and 2 Gynaecology and Reproductive Science, Università degli Studi di Padova, Padova, Italy; 3 Department of Pathology, Hospital Universitario San Cecilio, Granada, Spain; and 4 Department of Pathology, Hospital Virgen del Camino, Pamplona, Spain
Requests for reprints: Francisco F. Nogales, Departamento de Anatomía Patológica, Facultad de Medicina, 18012 Granada, Spain. Phone: 34958243508; Fax: 34958243510; E-mail: fnogales{at}ugr.es.
| Abstract |
|---|
|
|
|---|
Experimental Design: Immunohistochemical analysis of 53 instances of morular metaplasia comprising 1 cyclic endometrium and 52 endometrioid lesions associated with focal glandular complexity corresponding to 9 polyps, 4 atypical polypoid adenomyomas, 24 complex endometrial hyperplasias (18 with and 6 without atypia), 12 grade 1 endometrioid adenocarcinomas in early clinical stages of both uterus and ovary, and three ovarian adenofibromas. Immunohistochemistry in paraffin sections was done for CD10, ß-catenin, estrogen and progesterone receptors, and cytokeratins 5-6, 7, 8, 13, 18, 19, 20, and 34ß-E12.
Results: Morules were negative for estrogen and progesterone receptors and had ß-cateninpositive nuclei. Cytokeratins 8, 18, 19 were positive; cytokeratins 7 and 20 were negative; and cytokeratins 5-6, 13, and 34ß-E12 were weakly positive. All cases revealed strongly positive membranous CD10 staining in morules, which was absent in glands. CD10 positivity allowed easy identification of morules at low power in various types of surgical specimens and in curettings. CD10 also highlighted early morular metaplasia in glandular epithelium. In cases associated with squamous, keratinizing metaplasia, CD10 discriminated between both types of metaplasia.
Conclusions: CD10 staining represents a useful marker of morules in endometrioid neoplasms of the female genital tract, permitting identification of lesions usually associated with an attenuated malignancy. Considering the immunohistochemical and genetic similarities of morules in tumors of different organs, it is likely that this marker may be also useful to diagnose morular metaplasia in similar neoplasms of extragenital locations.
In both eutopic and ectopic endometrioid tissues, morules are almost invariably associated with the glandular architectural complexity of premalignant and low-grade glandular malignant lesions but are absent in high-grade ones. Thus, it can be said that the presence of morules in the endometrium relates well with an attenuated malignancy.
This study deals with the immunohistochemical findings in a series of 53 cases of morular metaplasia from both uterus and ovary, showing for the first time that CD10 staining is a marker of morules, allowing its easy identification in various endometrioid lesions.
| Materials and Methods |
|---|
|
|
|---|
|
Immunohistochemistry according to standard procedures was done using the antibodies listed in Table 2 .
|
| Results |
|---|
|
|
|---|
Microscopy
Typical morules were rounded, well-circumscribed aggregations of uniform, oval, or spindle-shaped cells with regular nuclei and eosinophilic cytoplasm. They developed within the glandular epithelium of endometrial glands (Fig. 1A
) or in the glandular spaces of endometrioid ovarian tumors (Fig. 1B), protruding into glands and eventually plugging the entire lumen. In some cases, morules were formed by sheets of cells that obliterated the supporting glands creating discrete stromal nodules. Except for one case, where they occurred in an otherwise typical proliferative endometrium, they were always associated with focal complex glandular lesions (Fig. 1C) with (Fig. 1D) or without atypia (Fig. 1E). In the ovary, endometrioid adenofibromas with glandular complexity were almost invariably associated with morules.
|
Associated keratinization only occurred in 11 of the 53 cases, of which nine corresponded to squamous differentiation within the morular nodules (Fig. 1F) and two to areas of squamous metaplasia elsewhere in the specimen but not within the morules. They corresponded to one polyp, one atypical polypoid adenomyoma, one complex hyperplasia without atypia, six complex atypical hyperplasias, and two adenocarcinomas. Two progestin-treated cases of atypical hyperplasia with repeat biopsies revealed, on follow-up biopsies, features of squamous differentiation within the morules that were not evident in the morular components of initial biopsy.
The morules present in the two cases of synchronous endometrioid tumors of endometrium and ovary revealed no morphologic differences.
Immunohistochemistry
The immunohistochemical findings in morules are summarized in Table 2. Morules were negative for estrogen and progesterone receptors and had ß-catenin positive nuclei, and their cytokeratin profile was positive for cytokeratins 8, 18, and 19; negative for cytokeratins 7 and 20; and weakly positive for cytokeratins 5-6, 13, and 34ß-E12.
All of our cases revealed strongly positive membranous CD10 staining in the morules (Fig. 2A ), which contrasted with the uniformly negative glandular epithelium. CD10 positivity allowed morular identification at low-power magnification (Fig. 2B), and the stain was evident even in some moderately autolytic endometria of hysterectomy specimens. Positivity of endometrial stroma provided an internal staining control. CD10 proved to be a consistent marker in various stages of development of morules. Early morular growth within the glandular epithelium was shown by focal CD10 positivity even when it was not apparent in the H&E stain (Fig. 2C and D). In morules of ovarian endometrioid lesions, CD10 staining pattern was identical.
|
The other constant immunohistochemical feature was the intense nuclear ß-catenin positivity of morules, which differentiated them from the adjacent glandular epithelium, which only had a membranous expression (Fig. 3A ). Early intraepithelial morular change was also easily shown in focal areas of glandular epithelium by nuclear ß-catenin and correlated well with CD10-positive focal glandular areas. Nuclear expression of ß-catenin was absent in the surrounding endometrial stroma.
|
Optically clear nuclei were present in seven cases and had large but well-delineated biotin-positive vacuolae that were easily differentiated from the diffuse nuclear staining present in both estrogen and progesterone receptors and ß-catenin.
The cytokeratin staining profile revealed differences between morules and glandular epithelium; whereas cytokeratin 5-6, 8, 13, 18, 19, and 20 staining patterns were similar, cytokeratin 7 staining was absent in morules and 34ß-E12 was weakly positive in morules but negative in endometrioid glands (Table 2).
The cases with keratinized morules and the five endometrioid lesions with keratinizing squamous differentiation used as controls revealed a different staining pattern from that of the morules (Table 3 ).
|
The other patient was a 38-year-old with a clinical history of infertility who underwent surgery to remove an ovarian endometrioid adenofibroma associated with a complex glandular focus with morules. Subsequently, after an initial diagnosis of focal endometrial atypical hyperplasia with morules, she had five endometrial curettages over a period of 4 years. The successive biopsies taken after treatment with progestins showed a persistent lesion with a gradual increase of glandular complexity. Their morules developed squamous metaplasia and a correspondingly decreasing CD10 stain.
Patients with repeat biopsies. Three further patients had follow ups with repeat endometrial biopsy.
The first was a 50-year-old with an initial diagnosis of complex atypical hyperplasia with both morular and squamous keratinizing metaplasia. No progestogen treatment was given. Two further endometrial biopsies followed by a hysteroscopic endometrial ablation and finally hysterectomy revealed a persistent, morphologically similar lesion.
The second case was a 35-year-old infertile patient treated with progestogen. An initial endometrial biopsy showed a simple endometrial hyperplasia and progestin effect. A second endometrial biopsy 1 year later showed a focal complex atypical hyperplasia with morules. Seven months later, a curettage showed an increase in glandular complexity and some cribriform glands, which prompted a diagnosis of well-differentiated endometrioid adenocarcinoma with morular metaplasia. In the hysterectomy specimen, there was only a residual complex atypical hyperplasia with morules. Morules had become keratinized after treatment and had a loss of CD10 expression.
The third case corresponded to an infertile 21-year-old patient. A first endometrial curettage showed a complex atypical hyperplasia with extensive morular metaplasia. She was subsequently treated with progestins. Treatment was discontinued after 1 year, and she became pregnant and had a successful term delivery. She recommenced progestogen treatment and had repeat biopsies taken during a period of 7 years, which revealed persistence of glandular hyperplastic atypical lesions and morular metaplasia coexisting with squamous metaplasia.
| Discussion |
|---|
|
|
|---|
Morules are highly characteristic structures with immunohistochemical (9, 10) and genetic profiles that present marked differences with those of normal endometrium and endometrioid neoplasms (10, 11), a finding that suggests that they may represent a characteristic, independent type of neoplastic transformation frequently involving a mutation of ß-catenin but not of PTEN, K-ras, and microsatellite instability found in endometrioid neoplasms without a morular component (1113).
Although more frequent in the endometrium, morules also occur in low-grade glandular lesions of other anatomic sites (38), where they seem to display a similar immunohistochemical pattern of nuclear ß-catenin staining (14) and a genetic profile involving ß-catenin mutations (5). Due to their similar morphologic, immunohistochemical, and genetic features in different organs, morules seem to represent a stereotyped change sharing an aberrant nuclear ß-catenin accumulation probably secondary to ß-catenin mutation (11). This protein has been described as a component of the cadherin-mediated intercellular adhesion system (15, 16) and in association with Wnt signaling pathway, where it can play a role in cancer genesis (17). Morules seem to be usually associated with an attenuated malignancy (18), with longer cancer-specific survival rates (12).
The present article deals with a series of low-grade endometrioid lesions associated with morular metaplasia. All cases except one corresponded to complex glandular lesions with varying degrees of atypicality. Morules originated as intraepithelial glandular foci easily differentiated from the glandular epithelium by their eosinophilic cytoplasm. Expansile intraglandular growth eventually obliterated and atrophied the remaining gland perimeter, sometimes leading to an appearance of "stromal" morules. Central necrosis of morules occurred more frequently in the larger ones, probably due to ischemia. Biotin-rich, optically clear nuclei were present; a feature that has been reported in morules elsewhere and is associated with estrogen receptor-ß expression (14).
In the present series, morules occurred with a higher frequency in hyperplastic endometrial lesions corresponding to both complex hyperplasias with or without atypia. In ovarian and uterine neoplasms, such as polyps, atypical polypoid adenomyomas, ovarian adenofibromas, and well-differentiated endometrioid adenocarcinomas, they invariably occurred associated with architecturally complex glandular lesions.
All cases of uterine and ovarian endometrioid tumors were in early clinical stages, and none have either recurred or metastasized, thus showing an attenuated malignancy for these lesions. It is worth noting that in the two patients with synchronous endometrial and ovarian tumors, morular metaplasia occurred in both locations.
Cases where repeat biopsies were available revealed the persistence of lesions during progesterone treatment. Even in a case of atypical hyperplasia with an intervening pregnancy and several courses of progesterone treatment, morules were seen in six consecutive biopsies during a 7-year follow-up. In the other two progesterone-treated cases, morules remained unchanged except for the presence of intramorular keratinization after treatment.
Previous immunohistochemical studies in morules of endometrial lesions and other organs reveal a characteristic immunophenotype. Morular nuclei are invariably negative to estrogen and progesterone receptors (9, 19) and positive for ß-catenin (25, 912, 14, 18, 19), always contrasting with the well-differentiated glandular epithelium, which is positive to receptors and practically only exhibits membrane-bound ß-catenin. Recently, various neuroectodermal markers (9, 20) have been also shown to be positive in morules.
In gynecologic pathology, CD10 is a useful antibody that helps to show both endometrial stromal (21) and trophoblastic cells (22) and is also positive in tumors with a mesonephric component (22). This article reports for the first time the role of CD10 as a marker of the morular component of various endometrioid lesions. Its strong membranous staining differentiates its cells from those of the endometrial glandular epithelium, which is usually negative. Only occasional strong CD10 positivity has been reported in glands of endometrioid adenocarcinoma (23). However, a previous report failed to show its positivity in endometrial morules (19), and only in one recent article (24) was CD10 positivity shown in morules of colonic neoplasms. We have been able to show identical CD10 strong positivity in morules of well-differentiated fetal adenocarcinoma of the lung.5 CD10 permits identification of morules at low power in various types of surgical specimens and in curettings, particularly in fragmented aspirative biopsies. Positive CD10 highlights morules that can be easily differentiated from the stroma, which is also CD10 positive, both by their glandular relationships and their nuclear positivity of ß-catenin and negativity to steroid receptors. CD10 positivity in morules remains even in moderately autolytic endometrial lesions and necrotic debris. CD10 also shows early morular change in glandular epithelium, which also correlates well with the receptor and ß-catenin stains.
Keratinizing squamous metaplasia is frequently associated to endometrioid adenocarcinoma and must be differentiated from morular metaplasia in the 20% of cases where they may coexist. CD10 helps to differentiate between morular and keratinizing squamous metaplasia because the latter has a staining pattern similar to glandular epithelium.
In this article, we have shown that CD10 staining represents a useful marker of morules in endometrioid lesions of the female genital tract. We believe that the strong positivity of CD10 in morules permits the easy identification of lesions usually associated with an attenuated malignancy. Taking into account the immunohistochemical and genetic similarities of morules in tumors of different organs, it is likely that this marker may be also useful to facilitate the diagnosis of morular metaplasia in these extragenital locations.
| Footnotes |
|---|
5 F.F. Nogales, L. Gonçalves, and C. Buriticá, unpublished observation. ![]()
Received 2/20/06; revised 4/20/06; accepted 5/10/06.
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |