OBJECTIVE: To identify in cytology, high-grade squamous intraepithelial lesions with endocervical glandular extension in cases previously diagnosed as atypical glandular cells (AGC), analyse possible reasons for the diagnostic pitfall and document the frequency of glandular pathology coexisting with high-grade cervical intraepithelial lesion in histology. METHODS: Thirty-nine ThinPrep cervical smear (Pap) tests reported as AGC of undetermined significance and showing high-grade lesions on histology [cervical intraepithelial neoplasia (CIN) 2 or 3, endometrial or extrauterine adenocarcinoma] were reviewed retrospectively to identify the cases of high-grade squamous intraepithelial lesion with endocervical glandular extension, using the Bethesda 2001 system. Cyto-histological correlation was performed. RESULTS: A high frequency of diverse glandular pathologies coexisted with high-grade cervical intraepithelial lesions on histology. This included endocervical glandular extension in 63%, benign glandular pathology in 33% and pre-neoplastic or malignant glandular pathology (endocervical glandular dysplasia, adenocarcinoma in situ and metastatic breast carcinoma) in 17% cases. On cytology, the sensitivity was 40%, specificity was 80% and positive predictive value was 86% for endocervical gland extension in high-grade squamous intraepithelial lesions. CONCLUSIONS: Special efforts to recognize endocervical glandular extension in high-grade squamous intraepithelial lesions and glandular neoplasia coexisting with squamous intraepithelial lesions from the heterogeneous category of AGC can contribute to increasing the diagnostic accuracy. The identification of endocervical glandular extension on cervical cytology would alert the gynaecologist to perform a thorough assessment of the endocervix during colposcopy. This could also help to decide on the need to perform deeper conization rather than loop electrosurgical excision procedure to ensure negative margins when colposcopic biopsy shows CIN 2 or 3.
OBJECTIVE: To identify in cytology, high-grade squamous intraepithelial lesions with endocervical glandular extension in cases previously diagnosed as atypical glandular cells (AGC), analyse possible reasons for the diagnostic pitfall and document the frequency of glandular pathology coexisting with high-grade cervical intraepithelial lesion in histology. METHODS: Thirty-nine ThinPrep cervical smear (Pap) tests reported as AGC of undetermined significance and showing high-grade lesions on histology [cervical intraepithelial neoplasia (CIN) 2 or 3, endometrial or extrauterine adenocarcinoma] were reviewed retrospectively to identify the cases of high-grade squamous intraepithelial lesion with endocervical glandular extension, using the Bethesda 2001 system. Cyto-histological correlation was performed. RESULTS: A high frequency of diverse glandular pathologies coexisted with high-grade cervical intraepithelial lesions on histology. This included endocervical glandular extension in 63%, benign glandular pathology in 33% and pre-neoplastic or malignant glandular pathology (endocervical glandular dysplasia, adenocarcinoma in situ and metastatic breast carcinoma) in 17% cases. On cytology, the sensitivity was 40%, specificity was 80% and positive predictive value was 86% for endocervical gland extension in high-grade squamous intraepithelial lesions. CONCLUSIONS: Special efforts to recognize endocervical glandular extension in high-grade squamous intraepithelial lesions and glandular neoplasia coexisting with squamous intraepithelial lesions from the heterogeneous category of AGC can contribute to increasing the diagnostic accuracy. The identification of endocervical glandular extension on cervical cytology would alert the gynaecologist to perform a thorough assessment of the endocervix during colposcopy. This could also help to decide on the need to perform deeper conization rather than loop electrosurgical excision procedure to ensure negative margins when colposcopic biopsy shows CIN 2 or 3.