Use Of Epithelial-Specific Antigen For Cytological Diagnosis Of Glandular Lesions In The Uterine Cervix

To the Editor: It has been reported that endocervical adenocarcinoma accounts for 5–20% of all cervical carcinomas and its incidence has been increasing in recent years. Endocervical adenocarcinoma displays less radiosensitivity than squamous cell carcinoma and effective chemotherapy has not been established. Therefore, endocervical adenocarcinoma currently has a poor prognosis, although it is hoped that early detection and early treatment may result in prognostic improvement. However, the accuracy of cytological diagnosis for detecting endocervical adenocarcinoma is not always high due to problems with the cellular sampling method and with screening. Regarding cellular sampling, if the site from which the endocervical lesion cannot be observed by colposcopy, it can be difficult to obtain sufficient amounts of tumor cells. Regarding problems with screening, concomitant squamous epithelial lesions, observed in 40–60% of patients with endocervical adenocarcinoma, making it difficult to screening for adenocarcinoma cells. Additionally, the false-negative rate of cytological diagnosis for endocervical adenocarcinoma is reported to range from 12.8% to 55.0%. This data suggests that it is difficult to use cytological diagnosis for endocervical glandular lesion. It has been reported that the expression rate of epithelialspecific antigen (ESA) is high in histologically detected endocervical adenocarcinoma. The present study was performed to determine whether or not expression of ESA is specific for endocervical lesions in cytology specimens and histological specimens. The subjects of this study were a total of 61 patients, who underwent both cytological and histological diagnosis (biopsy and excised specimens) almost simultaneously at our hospital from 1990 to 2007. They included 20 patients with endocervical adenocarcinoma (EA), 10 patients with adenocarcinoma in situ (AIS), 11 patients with carcinoma in situ (CIS), and 20 patients with invasive squamous cell carcinoma (SCC). Epithelial-specific antigen (VU1D9, 1:100; Novocastra Laboratories, Newcastle upon Tyne, UK) was detected by the indirect immunohistochemical method (MAX-PO MALT; Nichirei Biosciences, Tokyo, Japan) using HE-stained and Papanicolaou-stained (Pap-stained) and decolorized specimens. When histological specimens were investigated, antigen activation was performed by pretreatment with 0.1% trypsin at 37°C for 30 min. In both histology specimens and cytology specimens, endogenous peroxidases were inactivated by incubation with 3% H2O2 for 30 min. Both the primary antibody and the secondary antibody were reacted with the specimens for 30 min at room temperature, after which color was developed with DAB. The percentage of ESA-positive cells among 500 tumor cells was calculated in both histology specimens and cytology specimens and was classified as follows: 0% = − ; ≤ 30% = 1+ ; 30–70% = 2+ ; ≥ 70% = 3+ . Evaluation of the ESA expression showed the positive reaction to the cell membrane or cytoplasm of epithelial cells. For statistical analysis, Student’s t-test was employed and P < 0.05 was considered to be statistically significant. Expression of ESA for histological findings, normal squamous cells were generally negative for ESA, although some of the basal cells were positive. Endocervical cells were mainly negative for ESA, though the basolateral membranes were partially positive (Fig. 1a). Squamous metaplasia partially showed weak expression of ESA. Expression of ESA in endocervical lesions was seen in 85.0% of EA lesions (17/20 patients) (Fig. 1b); 100.0% of AIS lesions (10/10 patients) (Fig. 1c); 18.2% of CIS lesions (2/11 patients); and 45.0% of SCC lesions (9/20 patients). Thus, expression of ESA was detected in all patients with AIS and nearly all patients with EA (sensitivity of 90.0% and specificity of 64.5%). The level of ESA expression in patients with EA was mainly 2+ (1+ in 25.0% of the patients; 2+ in 45.0%; and 3+ in 15.0%), although there were individual differences. All of the patients with AIS had ESA-positive tumors and strong expression was frequent (3+ in 70%). On the other hand, most of the patients with CIS had ESA-negative tumors (81.8%) (Fig. 1d), but 40% of the patients with SCC had ESA-positive tumors (1+ in 25.0% and 2+ in 15.0%). In cytological specimens, normal squamous cells, squamous metaplastic cells and endocervical cells were negative for ESA or slight cytoplasmic expression in endocervical cells (Fig. 2a). Expression of ESA in endocervical lesion was seen in 100.0% of EA lesions (20/20 patients) (Fig. 2b) and 100.0% of AIS lesions (10/10 patients) (Fig. 2c) while ESA was only positive in 18.2% of CIS lesions (2/11 patients) (Fig. 2d) and 30.0% of SCC lesions (6/20 patients). Thus, all of the patients with EA or AIS were ESApositive, whereas the ESA expression rate was low in squamous lesions (sensitivity of 100.0% and specificity of 74.2%). The level of ESA expression was strong in the patients with EA or AIS (3+ in 100.0% of the patients with EA and 80.0% of the patients with AIS) (Fig. 2a, b). On the other hand, weak ESA expression was seen in patients with CIS or SCC (1+ in doi:10.1111/pin.12379 bs_bs_banner