Expression of epithelial membrane and 3-fucosyl-N-acetyllactosamine antigens in cervix uteri with particular reference to adenocarcinoma in situ

Factors predicting disease outcome in early stage adenocarcinoma of the uterine cervix

CONTEXT

Adenocarcinoma (AC) and adenosquamous carcinoma (ASC) comprise the second principal histological types of cervical carcinoma. As compared with the squamous cell cancer (SCC), these lesions are far less frequent, and their epidemiology, natural history and prognostic determinants are less well understood.

OBJECTIVE

Patients with an early stage AC of the uterine cervix diagnosed in our clinic were subjected to detailed analysis for the prognostic determinants.

STUDY SUBJECTS

A series of 94 women with early stage (adenocarcinoma in situ (AIS) to IIB) cervical ACs or ASCs diagnosed and treated in our department during 1995-1999 and subsequently followed-up for a mean of 43.1 +/- 16.2 (S.D.) months.

MAIN OUTCOME MEASURES

Patients were examined by colposcopy, Papanicolaou (PAP) smear and biopsy. The stage of the disease (FIGO) and tumour histology in operative specimens were recorded, and univariate (Kaplan-Meier) and multivariate survival analysis (Cox) were run to explore the factors predicting disease outcome.

RESULTS

Mean age of the women was 44.2 +/- 2.5 (S.D.) years (range 24-81 years), which is significantly (P=0.000) lower than that (49.9 +/- 14.2) of 464 SCC patients in our material. Minority of the women (38.2%) reported any clinical symptoms, but these correlated with the stage (P=0.041). Screening history was acceptable (i.e. screening interval 3 to 4 years) in 56 women, whereas 28 (29.8%) had no previous PAP smear taken. Interpretation errors were established in 17 (23.6%) and sampling errors in 6 (8.3%) of the 72 smears available for re-screening. No colposcopic lesions were found in 29 (30.9%) women. Follow-up data were available from 72 patients, of whom the disease progressed in four (one died), whereas 68 patients are alive and well at the moment. Patient's age (P=0.000), screening history (P=0.0127), FIGO stage (P=0.001), mode of therapy (P=0.0187), and presence of co-existent squamous cell lesions (P=0.0184) were significant prognostic indicators in univariate survival analysis. Cox's multivariate survival analysis disclosed FIGO stage (P=0.001) and screening history (P=0.006) as the only significant independent predictors of the disease outcome.

CONCLUSIONS

The present data emphasise the importance of early cervical AC as a disease of younger women, making early detection of its precursors (AIS) by regular PAP smear screening mandatory in prevention of disease progression. This can only be achieved by increasing the sensitivity of the PAP smear in detecting abnormal glandular cells in asymptomatic women.

Distribution of epithelial-specific antigen in uterine cervix with endocervical glandular dysplasia

The natural history of cervical adenocarcinoma is largely unknown, and whether endocervical glandular dysplasia is a precursor lesion to malignant tumor formation is a controversial issue. The aim of this study was to clarify the relationship of endocervical glandular dysplasia to the uterine endocervical adenocarcinoma. Thirty-one cases of glandular abnormalities of the uterine cervix were identified histologically from January 1984 to April 1992. These included 11 cases of endocervical glandular dysplasia, 5 cases of adenocarcinoma in situ, 4 cases of microinvasive adenocarcinoma, and 11 cases of invasive adenocarcinoma. The immunohistochemical localization of epithelial-specific antigen (ESA) was examined in these cases of endocervical glandular dysplasia and related lesions, and was compared to 10 normal endocervical specimens. ESA immunoreactivities were usually present in the basolateral membrane of endocervical cells in the normal endocervix. However, the expressions of ESA were increasing from the basolateral membrane to the diffuse cytoplasmic membrane, along with the malignant transformation of the endocervical cells, in 6 of 11 cases with glandular dysplasia, in 9 of 9 cases with in situ and microinvasive adenocarcinomas, and in 11 of 11 cases with invasive adenocarcinomas. This finding indicates that ESA is a useful marker for endocervical glandular dysplasia and related lesions with malignant transformation, and suggests that endocervical glandular dysplasia may be a precursor lesion to uterine cervical adenocarcinomas.

Endocervical carcinoma and precursor lesions: c-myc expression and the demonstration of field changes

The monoclonal antibody Myc 1-6E10 was used in an immunocytochemical technique to examine the expression of the c-myc oncogene in normal endocervices and those with glandular intraepithelial neoplasia and invasive malignancy. Eleven of 14 normal endocervical biopsy specimens did not express the gene, while three showed nuclear, or light basal cytoplasmic localisation of the gene product, or both. All but one of 14 cases of low and high grade glandular intraepithelial neoplasia, and all 17 cases of invasive adenocarcinoma, showed a pan-cellular pattern of immunostaining. Of considerable additional interest was the demonstration of field changes in morphologically normal glandular epithelium in those biopsy specimens with concurrent cervical glandular intraepithelial neoplasia or adenocarcinoma. This was manifest as increased concentrations of myc proteins compared with normal tissues. These results further support the role of the c-myc gene in oncogenesis, and in the light of field changes, suggest possible difficulties in the clinical management of this group of patients.

Nucleolar organiser regions in adenocarcinoma in situ and invasive adenocarcinoma of the cervix

Silver binding nucleolar regions (AgNORs) were evaluated in normal endocervix, adenocarcinoma, and its potential precursor, adenocarcinoma in situ (AIS), in an attempt to increase an understanding of the natural history of cervical adenocarcinoma and to identify a marker for abnormal endocervical (atypical glandular) cells which could aid diagnosis and follow up of endocervical lesions. For every 50 cells the mean AgNOR counts were as follows: normal endocervical cells (n = 15) 79.8 (95% Cl 68-91); AIS (n = 20) 200.7 (95% Cl 182-219); and invasive adenocarcinoma (n = 30) 299 (271-328). There was no overlap between the groups of normal endocervical cells and invasive adenocarcinoma, but there was significant overlap between cases of invasive adenocarcinoma and carcinoma in situ. In six out of 17 cases with AIS, NOR count in adjacent morphologically normal glandular cells ("internal" controls) was increased when compared with the "external" (normal endocervical) control group. This suggests the presence of wider field changes not previously identified using routine histological methods. The findings suggest that AIS is a potential premalignant precursor of invasive adenocarcinoma, but that assessment of NORs is of no practical use in discriminating between the histological types of cervical carcinoma.